Impact of organized activities on mental health in children and adolescents: An umbrella review

Highlights • This umbrella review included six systematic reviews of lower quality.• A small positive impact of organized sport activities on youth mental health was found.• Only two systematic reviews that studied organized non-sport activities could be included.• Mixed results were found for impact of organized non-sport activities on youth mental health.• Mediation or confounding through physical activity needs further study.


Introduction
Mental health problems are a leading cause of health-related disability among children and adolescents (Kieling et al., 2011;Keyes, 2002). Worldwide, around 10-20% of all children and adolescents experience mental health problems (Kessler et al., 2007). Mental health problems during childhood or adolescence can also have implications in adulthood such as mental health problems or employment difficulties (Copeland et al., 2015;Fergusson et al., 2007;Ning et al., 2020). However, according to the World Health Organization (WHO) mental health is more than the absence of mental health problems. It is "a state of well-being in which the individual realizes his or her own abilities, can cope with the normal stresses of life, can work productively and is able to make a contribution to the community" (World Health Organization. Promoting mental health: concepts, emerging evidence, practice: summary report. Geneva;, 2004). In a similar vein, Keyes suggests that absence of mental health (i.e. languishing), is not similar to good mental health (i.e. flourishing) (Galderisi et al., 2015;Keyes, 2002Keyes, , 2005Keyes, , 2007Westerhof and Keyes, 2010;King et al., 2014). Exploring possible preventive factors for reducing mental health problems and for promoting good mental health is thus important.
One possible factor for reducing mental health problems or promoting good mental health is participation in organized activities, whether sport or non-sport (Bohnert et al., 2010;Lerner et al., 2012).
The positive youth development theory, grounded in the socioecological systems theory, postulated that sport and non-sport organized activities may offer opportunities for children and adolescents to develop relationships, engage in activities that increase their confidence, competence, character, caring and connectedness (Agans et al., 2016;Bronfenbrenner, 2003). Consequentially it is hypothesized that they are at lower risk for academic, psychological, social and behavioral problems (Bohnert et al., 2010;Lerner et al., 2012). This may depend on the type, breadth, intensity and duration of the activities (Bohnert et al., 2010).
Organized sport and non-sport activities can be defined as activities that are structured, supervised by adults, emphasize skill building, are generally voluntary, have regular scheduled meetings and are not part of the school curriculum (Bohnert et al., 2010;Mahoney et al., 2005a,b). Examples of organized activities include but are not limited to sport, arts, music and community programs (Bohnert et al., 2010;Mahoney et al., 2005a,b). Features of organized activities that have been found to improve mental health are: safe and appropriate peer interactions, structure and adult supervision, forming of supportive relationships with peers and adults, emphasis on inclusion and a sense of belonging, emphasis on positive social norms, support of efficacy and mattering and skill-building (Mahoney et al.,). Organized sport activities also includes physical activity as an additional feature that may improve the mental health of children and adolescents (Biddle and Asare, 2011). Organized non-sport activities may include physical activity but not always (e.g. scouting or dance). Local policies can influence and encourage participation in organized sport and non-sport activities and its determinants.
Several studies examined the impact of various types of organized sport and non-sport activities on aspects of child and adolescent mental health. These suggest a possible beneficial impact on mental health such as behavioral outcomes, self-esteem and self-confidence (Bungay and Vella-Burrows, 2013;Zarobe and Bungay, 2017). Contrary, some studies have observed harmful consequences of organized activities, such as risk behavior and bullying (Badura et al., 2017;Fredricks and Eccles, 2008;Matjasko et al., 2019;Randall and Bohnert, 2012) In two of these studies this depended on the amount of time that was spend in the organized activities (Matjasko et al., 2019;Randall and Bohnert, 2012) () As far as we know, there currently is no overview aggregating the available evidence from systematic reviews on the impact of participation in organized sport and non-sport activities on child and adolescent mental health, while such an overview would be highly relevant for policymakers in designing more effective preventive youth policies. Thus, the aim of this study is providing an overview of the evidence of the impact of organized sport and non-sport activities on childhood and adolescent mental health outcomes from a public health perspective based on an umbrella review of published systematic reviews.

Methods
A protocol prospectively registered within PROSPERO (CRD42020213597, available via https://www.crd.york.ac.uk/prosp ero/display_record.php?ID=CRD42020213597) on November 9th 2020. The protocol was developed in accordance with the Preferred Reporting Items for systematic reviews and meta-analyses Protocol (PRISMA-P) (Moher et al., 2015). This umbrella review adheres to the PRISMA statement and used guidance from Arotomatis et al. (Aromataris et al., 2015;Page et al., 2020).

Eligibility
Peer-reviewed systematic reviews with and without meta-analyses were considered eligible (Aromataris et al., 2015). Other types of reviews (i.e. narrative or rapid reviews) were excluded as we aim to summarize studies with the highest level of evidence (Ioannidis, 2009). Systematic reviews containing observational study designs (i.e. casecontrol, cohort, cross-sectional) and trial designs in any form including pilot studies were included. Qualitative designs were considered not eligible. Systematic reviews containing both quantitative and qualitative designs were included.
Systematic reviews containing studies with children and adolescents with a mean age between 0 and 21 years old were included (Hardin et al., 2017). Systematic reviews in which any participant was aged ≥ 25 years old without sub-analysis for participants aged < 25 years old with a mean age between 0 and 21 years old, were excluded. Systematic reviews on general populations were included, as well as reviews with sub-analyses on general populations. Systematic reviews that included studies on fully clinical or at-risk populations (e.g. attention deficit hyperactivity disorder (ADHD), traumatic experiences) were excluded as we aimed to study the impact of organized activities on mental health from a public health perspective.
Systematic reviews containing organized sport or non-sport activities were included. For this umbrella-review a definition based upon the definition postulated by Bohnert et al., was used (Bohnert et al., 2010;National Research Council and Institute of Medicine, 2002). Their definition is: "Organized activities is a blanket term that refers to a broad range of adult-sponsored activities that fall outside the regular school curriculum and include diverse contexts such as school-based extracurricular activities, community organizations, and youth development programs. Despite the differences in focus, organized activities share several common features. The activities are generally voluntary, hold regularly scheduled meetings, are supervised by adults, include other participants, are organized around particular competencies, and tend to be rule-based." (Bohnert et al., 2010;National Research Council and Institute of Medicine, 2002) In this umbrella review we broaden this definition and include organized activities without other participants (e.g. individual arts or music lessons, individual resistance training), organized activities that are not rule-based. Organized activities should be provided by a volunteering-or non-volunteering party (i.e. not organized by children or adolescents themselves)." The control group or comparator condition is formed by children and adolescents not exposed to organized sport activities and/or non-sport activities (e.g. non-organized sport or non-sport activities, no sport or non-sport activities).
There were no limitations regarding country. Systematic reviews were included if they included organized activities occurring in an extracurricular (after-school) or community (e.g. clubs, community centers) setting corresponding with our definition. Systematic reviews focusing on sport or non-sport activities that took place within schoolcurricula or clinical settings were excluded.
The outcomes of interest were indicators of me mental health as defined by the WHO and by Keyes (Galderisi et al., 2015;Keyes, 2002Keyes, , 2005Keyes, , 2007Westerhof and Keyes, 2010;King et al., 2014;World Health Organization, 2004). The WHO defined mental health as follows: "mental health is more than the absence of mental health problems. It is a state of well-being in which the individual realizes his or her own abilities, can cope with the normal stresses of life, can work productively and is able to make a contribution to the community" (World Health Organization, 2004). The definition of mental health by Keyes consists of three aspects: emotional well-being, psychological well-being and social well-being. According to Keyes absence of mental health problems (i.e. languishing) is not similar to good mental health (i.e. flourishing) and mental health should be seen as a continuum (Galderisi et al., 2015;Keyes, 2002Keyes, , 2005Keyes, , 2007Westerhof and Keyes, 2010;King et al., 2014). In line with these definitions we did not only include all (aspects of) mental health problems as indicators of mental health and thus as outcomes of interest (e.g. anxiety, depression, ADHD and other mental health problems), but also aspects of mental well-being (e.g. self-esteem, efficacy, self-worth). No preferred outcome measure was formulated a priori. Systematic reviews that did not report on at least one aspect of mental health were excluded. Systematic reviews published in English were included. Systematic reviews in other languages were excluded because of language barriers of the authors.

Search strategy
A systematic literature search was conducted in five databases (Embase.com, MEDLINE via Ovid, Web of Science core collection (See Appendix A for our core collection), CINAHL via EBSCOhost and Psy-cINFO via Ovid) from inception to March 25th 2021 (date last searched). The search strategy was developed by a Medical Librarian (WMB) and combined thesaurus terms as well as terms in title abstract for three elements: sport or participation, mental health or behavior, children or adolescents, and was limited to systematic reviews and meta-analysis. The syntax and thesaurus terms of the search strategy were adapted to each distinct database. In the search strategy, no language or date limits were applied. Appendix A includes the full search strategy. References of relevant reviews were screened for other relevant systematic reviews.

Selection process
Endnote X9 was used for the selection process. Duplicates of records were retrieved and removed using the method described by Bramer et al. (Bramer et al., 2016) Two independent reviewers (MB and MS) performed title and abstract screening to identify eligible reviews and subsequently performed full-text screening. Disagreements at both stages were resolved through discussion until consensus was reached, and, if necessary, resolved by consulting a third independent reviewer (WJ).

Data extraction
Data were extracted independently by two reviewers (MB and MS). Extracted information included: first author, year of publication, included languages of primary studies, objective, eligibility criteria, if it included a meta-analysis, number and type of primary studies, sample (size, age and sex), number of databases searched, range of publication date, instrument for quality appraisal and quality rating, intervention (s)/phenomena of interest, outcome(s) and outcome measure(s), measurement instruments and funding.

Overlap
The corrected covered area (CCA) was used to calculate the amount of overlap of primary studies included in the systematic reviews (Pieper et al., 2014). The CCA is calculated by dividing the frequency of repeated occurrences of index studies (first occurrence of primary study) in other reviews by the product of the number of index studies and the number of reviews, minus the number of reviews. The CCA can be represented as a percentage between 0 and 100%. A CCA of 0-5% is considered slight overlap, a CCA of 6-10% is considered moderate overlap, a CCA of 11-15% is considered high overlap and a CCA > 15 is considered very high overlap (Pieper et al., 2014). For the calculation, see Appendix B.

Risk of bias
Two reviewers (MB and MS) independently assessed risk of bias of the included systematic reviews using the A Measurement Tool to Assess Systematic Reviews (AMSTAR-2) (Shea et al., 2017). Discrepancies were resolved through discussion until consensus was reached and if necessary by consulting a third independent reviewer (WJ). The AMSTAR-2 consists of sixteen items. Seven items are considered critical, these are: a priori protocol, adequate search strategy, providing justification for excluded studies, appropriate risk of bias assessment, appropriate statistical methods in meta-analysis, accounting for risk of bias when interpreting results and publication bias. Three items concern meta-analytical methods and are not applicable for systematic reviews without meta-analysis. The AMSTAR-2 rates systematic reviews as: critically low (more than one critical weakness with or without noncritical weaknesses), low (one critical weakness with or without noncritical weaknesses), moderate (more than one non-critical weakness) or high quality (no or one non-critical weakness) (Shea et al., 2017).

Analysis
Because of the broad scope of organized activities and mental health outcomes included in this umbrella review performing a meta-analysis was not possible. Results were narratively (qualitative and quantitative) synthesized. Results were grouped by type of organized activities (i.e. sport, non-sport or both) and further subdivided by type of organized sport activity. As meta-analyses used different effect sizes or measures of association, we converted all reported effect sizes and measures of association to the Cohen's d, for comparison purposes (Fusar-Poli and Radua, 2018;Sanchez-Meca et al., 2003). Formulas for these conversions are reported in Appendix C. For the summary of findings, quality of evidence per determinant was assessed by a self-developed decision scheme including: meta-analysis, number of primary studies, significance, direction, magnitude and imprecision. Scores were: no indication, mixed findings, insufficient evidence, there is an indication or high certainty. A self-developed scheme was used as suitable schemes for umbrella reviews are lacking (Appendix D).

Results
After deduplication, 833 records remained. After all rounds of screening six systematic reviews were considered eligible (Cairns et al., 2014;Collins et al., 2019;Eime et al., 2013;Evans et al., 2017;Zuckerman et al., 2020;Panza et al., 2020). Fig. 1 describes the results of the search and study selection process. Appendix E includes references (n = 85) of all studies excluded after full-text screening subdivided by reason for exclusion. Table 1 shows the characteristics of the six included systematic reviews. Systematic reviews were published between 2013 and 2020 (Cairns et al., 2014;Collins et al., 2019;Eime et al., 2013;Evans et al., 2017;Zuckerman et al., 2020;Panza et al., 2020). Four systematic reviews included meta-analyses (Cairns et al., 2014;Collins et al., 2019;Eime et al., 2013;Zuckerman et al., 2020). Sample sizes of the systematic reviews ranged from 460 to 234,503 participants. Databases searched ranged from 1 to 13. Included primary studies ranged from 7 to 113 studies. Most primary studies were from North America, followed by Europe and Australia. Five primary studies were from Asia and two from Africa. Primary studies were published between 1988 and 2020.
All six systematic reviews reported on organized sport activities and two also reported on organized non-sport activities (Cairns et al., 2014;Eime et al., 2013). Five systematic reviews examined individual and team-based sport (Cairns et al., 2014;Eime et al., 2013;Evans et al., 2017;Zuckerman et al., 2020;Panza et al., 2020). Three systematic reviews examined level of sport involvement (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020). Two systematic reviews focused on extracurricular school and community non-sport activities and sport (Eime et al., 2013;Evans et al., 2017). One examined resistance training (Collins et al., 2019). One examined organized non-sport activities (Cairns et al., 2014). One examined (school) club sport and nonspecified sport (Eime et al., 2013). One examined non-specified sport. Table 1 shows all mental health outcomes studied in the included systematic reviews. Most studied mental health outcomes were (aspects of) mental health problems such as depressive symptoms and anxiety symptoms (Cairns et al., 2014;Eime et al., 2013;Zuckerman et al., 2020;Panza et al., 2020). Less studied were aspects of mental well-being such as development of positive assets, self-esteem, self-worth and selfconcept (Collins et al., 2019;Eime et al., 2013;Evans et al., 2017).
In total 17 studies out of 118 relevant primary studies were reported in multiple systematic reviews (14.4%). Three primary studies were reported thrice in the included systematic reviews (Gore et al., 2001;Sanders et al., 2000;Zarrett et al., 2009). The CCA amounting 3.6% indicates a slight overlap. Appendix B shows the citation matrix used to calculate the overlap. Table 2 reports the quality assessment. Five systematic reviews were identified as critically low (Cairns et al., 2014;Collins et al., 2019;Gore et al., 2001;Zuckerman et al., 2020;Panza et al., 2020). One systematic review was identified as moderate in quality (Eime et al., 2013). Common quality lowering items were lack of reporting funding in primary studies (6/6), lack of an a priori protocol (5/6), lack of a description of excluded studies (5/6). Report of funding of the primary studies is needed to assess possible bias such as changes in the design, analyses or conclusion in favor of the interests of the funder (McCrabb et al., 2021). An a priori protocol helps researchers conducting their review as it has been planned and reduces arbitrary decision-making (Moher et al., 2009). An a priori protocol also enables readers to identify deviations from the planned methods and selective outcome reporting (Moher et al., 2015). Justification for excluding studies is needed to examine the impact of their exclusion from the review (Shea et al., 2017). Table 3 and 4 provide meta-analysis and qualitative results, summarized findings and the quality of evidence.
Five systematic reviews reported results on team-based and individual sport participation (Cairns et al., 2014;Eime et al., 2013;Evans et al., 2017;Zuckerman et al., 2020;Panza et al., 2020). Three of these reported meta-analysis results of the impact of team-based and individual sport participation on mental health outcomes (Cairns et al., 2014;Eime et al., 2013;Panza et al., 2020). These mental health outcomes were either depressive symptoms, anxiety or a combination (Cairns et al., 2014;Eime et al., 2013;Panza et al., 2020). Summarized, a significant positive impact on mental health outcomes was found. However, the magnitude of the effect estimate of the mental health outcomes was negligible or small (i.e. reduced anxiety/depressive symptoms). The heterogeneity was either moderate or high. Three systematic reviews reported qualitatively on the impact of team-based sport on mental health outcomes (      Rating was as follows: high quality of review: No or one non-critical weakness, Moderate quality of review: More than one non-critical weakness, Low quality of review: One critical flaw with or without non-critical weaknesses, Critically low quality of review: More than one critical flaw with our without non-critical weaknesses. PICO = population, intervention/exposure, control/comparator, outcome; RoB = Risk of Bias. Zuckerman et al., 2020). The qualitative synthesis of team-based sport yielded mixed findings on mental health outcomes (Eime et al., 2013;Evans et al., 2017;Zuckerman et al., 2020). The mixed findings were due to some studies that found no positive or negative impact (null effect). Many different mental health outcomes were studied in the qualitative analysis (Eime et al., 2013;Evans et al., 2017;Zuckerman et al., 2020). Three systematic reviews reported results regarding level of sport involvement (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020). Level of sport involvement included frequency, a longer period/duration of sport involvement, intensity and involvement in sport at an early age (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020). One systematic review reported meta-analysis results of the frequency of sport involvement with depressive symptoms as mental health outcome (Panza et al., 2020). Results from this review showed that there was evidence for an impact on mental health outcomes but with a negligible magnitude of effect and high heterogeneity (Panza et al., 2020). Qualitatively, findings from three systematic reviews showed a positive impact of a higher frequency, greater intensity, a longer period/duration of sport involvement and involvement at an early age (i.e. during childhood) on mental health outcomes (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020).
Two systematic reviews reported results regarding extracurricular activities (Cairns et al., 2014;Eime et al., 2013). One systematic review reported meta-analysis results of non-sport extracurricular activities with depressive symptoms as mental health outcome (Cairns et al., 2014). Findings showed no impact of extracurricular non-sport activities on mental health outcomes as the meta-analysis yielded a non-significant result (Cairns et al., 2014). One systematic review reported qualitative results of extracurricular non-sport and sport school and community activities with mental health outcomes (Eime et al., 2013). Findings showed a positive impact on mental health outcomes such as, higher self-image, lower risk of emotional distress, better emotional regulation and psychosocial outcomes (Eime et al., 2013). This systematic review also reported that sport participation showed greater benefits than extracurricular non-sport activities and that a combination showed greatest benefits (Eime et al., 2013).
For resistance training, sport non-specified, (school) club sport, and other sport one systematic review reported results regarding mental health outcomes (Collins et al., 2019;Eime et al., 2013;Evans et al., 2017). Meta-analysis and qualitative findings of resistance training on mental health outcomes were mixed (Collins et al., 2019). Qualitative findings reported insufficient evidence for an impact on mental health outcomes from other sport and a positive impact on mental health outcomes from sport non-specified and (school) club sport. A small number of primary studies were included (Eime et al., 2013;Evans et al., 2017). Reported associations and effect sizes were transformed to Cohen's D effect sizes. The magnitude of Cohen's D was interpreted using Cohen's D conversion. Heterogeneity was assessed using the I 2 statistic. For interpretation, I 2 values of 25%, 50% and 75% were considered to indicate low, moderate and high heterogeneity. Summary of findings and quality of evidence is based on a self-developed decision scheme to assess the quality of evidence.

Table 4
Summarized findings of qualitative results and quality of evidence of included systematic reviews.
Author, year Primary studies used (n/ total) Results as extracted from systematic reviews Summary of findings and quality of evidence Sport organized activities Sport nonspecified Eime et al., 2013 4/30 There were findings that sport was associated with enhanced self-concept, lower rates of suicidal ideation (including thoughts and intentions), and with positive adjustment (e. g. social skills and selfesteem).
There is an indication of a positive impact on mental health outcomes by participating in no further specified sport activities.
Level of sport involvement (e.g. frequency, duration, intensity, early involvement) Eime et al., 2013 5/30 There were findings that greater sport participation was associated with lower risk of emotional distress and with lower levels of emotional and social problems. Also moderate sport participation was associated with lower depression scores. Greater participation in formal compared to informal sport was associated with lower levels of emotional and social problems. Greater frequency in sport participation led to better feelings of wellbeing compared to lower frequency. Total number of sport and years involved in sport was associated with better physical appearance and physical competence. Differences between competitive or noncompetitive sport were minimal. Evans et al., 2017 16/35 There were findings for an association of early sport involvement and amount of sport involvement with psychosocial outcomes (depression and selfesteem). There was insufficient evidence for amount of individual deliberate practice or specialization in sport due to limited research. Panza et al., 2020 3/29 There were findings that duration of sport participation may have a small inverse There were findings that support a positive effect of resistance training on some constructs of 'the self '.
There was a significant increase in total selfefficacy. No evidence for a positive effect of resistance training on self-concept.
There are mixed findings regarding the impact on mental health outcomes by participating in resistance training.
Team-based sport Eime et al., 2013 8/30 There were findings of mental health benefits (e.g. lower general risktaking, fewer mental and general health problems, positive associations with social acceptance and selfesteem and negative associations with depressive symptoms, social isolation and mood) by participation in team-based sport. There were also findings that it was protective against feelings of hopelessness and suicidality and that it increased life satisfaction.
There are mixed findings regarding the impact on mental health outcomes by participating in teambased sport.

Evans et al., 2017
14/35 There were findings of a positive association of participation in teambased sport to psychosocial outcomes (such as youth development experiences, moral reasoning, depression and self-esteem). Some studies reported null differences regarding depressive symptoms or anxiety. Zuckerman et al., 2020

23/34
The majority of studies supported a positive impact of team sport participation on many behavioral and psychological health outcomes. Additional studies found similarly positive effects such as less physical fighting.
(School) club sport Eime et al., 2013 5/30 There were findings of higher scores on social functioning and mental health by participating in school and club sport. There were also findings of an association with superior well-being (including being better adjusted) feeling less nervous or anxious, There is an indication of a positive impact on mental health outcomes by participating in (school) club sport.
(continued on next page)

Discussion
This umbrella review provides a detailed overview and shows that there may be a small positive impact on mental health in children and adolescents by participating in organized sport activities. Relatively much eligible research about organized sport activities and relatively less about organized non-sport activities with mental health in children and adolescents was found.
The objective of our umbrella review was to provide an overview of the evidence of the impact of organized sport and non-sport activities on child and adolescent mental health from a public health perspective. Fig. 2 shows the associations of interest for our umbrella review. Outside of the scope of our umbrella review and thus not studied are the association of participating in organized activities with physical activity and the association of physical activity with child and adolescent mental health. For the association of physical activity with child and adolescent mental health evidence of a positive impact was provided in the umbrella review of Biddle and Asare (Biddle and Asare, 2011). None of the systematic reviews included in our umbrella review had an objective similar to our own objective; all addressed different research questions. Cairns et al., conducted a systematic review to different modifiable risk and preventive factors associate with depression (Cairns et al., 2014). Individual and team-based sport and extracurricular activities were included as possible preventive factors and were discussed broadly (Cairns et al., 2014). Furthermore, Collins et al., and Zuckerman et al., both focused on one type of organized sport activities (i.e. resistance training and team-based sport) (Collins et al., 2019;Zuckerman et al., 2020). The other three systematic reviews focused on several types of organized sport activities and included aspects relevant for our objective such as distinguishing between different settings and patterns of involvement (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020). None of the systematic reviews particularly aimed to study organized non-sport activities but in two systematic reviews it was reported (Cairns et al., 2014;Eime et al., 2013). Overall, the systematic reviews reported a positive impact of organized sport activities on mental health although some negative or null results were found. Most studied mental health outcomes were (aspects of) mental health problems.
Six systematic reviews examined the impact of organized sport activities and there may be a small positive impact on child and adolescent mental health. However, effect sizes were negligible/small and heterogeneity between studies was high thus these findings need to be interpreted with caution. According to the positive youth development theory, children and adolescents can benefit from organized activities because they can develop relationships and engage in activities increasing their confidence, competence, character, caring and connectedness (Fraser-Thomas et al., 2005;Lerner et al., 2005;Agans et al., 2016;Benson et al., 2006;Holt et al., 2020;Lerner et al., 2012). Many types of organized sport activities are included in our umbrella review and findings regarding mental health were generally in the same direction. Unfortunately, we cannot disentangle the impact of There is insufficient evidence for an impact on mental health outcomes by participating in other categories of sport. .

Sport and non-sport organized activities
Extracurricular school and community non-sport activities and sport Eime et al., 2013 8/30 There were findings that structured activities (sport and non-sport) led to higher positive functioning. Children participating in sport and clubs had higher social skill scores compared to children who did not participate in outside-school activities. Participation in sport and non-sport organized activities led to the greatest youth development outcomes. Sport participation led to more developmental benefits than other types of extracurricular activities but the greatest benefits were seen for sport and nonsport extracurricular activities combined. Sport participation alone and in combination with nonsport activities was associated with better health outcomes, including higher healthy self-image, lower risk of emotional distress, suicidal There is an indication of a positive impact on mental health outcomes by participating in extracurricular and community non-sport activities and sport. Summary of findings and quality of evidence is based on a self-developed decision scheme to assess the quality of evidence.
participating in organized sport activities from physical activity. None of the reviews has reported on this possible confounding or mediation, and most primary studies did not adjust for physical activity. Moreover, we cannot exclude the possibility that good mental health leads to participation in organized sport activities (Sabiston et al., 2013). Further, selfselection or drop-out could possibly also lead to a seemingly better mental health among children and adolescents who participate in organized sport activities.
The positive impact of participating in organized sport activities was hypothesized to be dependent on the breadth, intensity, and duration of participation by Bohnert et al. (Bohnert et al., 2010) Findings from the systematic reviews were somewhat mixed. We found evidence that children and adolescents who participated more frequently, who participated with higher intensity or who participated with longer duration (e.g. started at an early age or participated for a long period), had better mental health compared to their peers (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020). However, the magnitude of the effect sizes were negligible (Eime et al., 2013;Evans et al., 2017;Panza et al., 2020).
We found no systematic review reporting exclusively on the impact of organized non-sport activities on mental health in children or adolescents. Only two systematic reviews reported on both organized sport and non-sport activities in relation to mental health among children and adolescents (Cairns et al., 2014;Eime et al., 2013). We are uncertain whether this is a gap in the literature due to scarcity of primary studies or that primary studies were not aggregated into systematic reviews. We found some systematic reviews that reported on organized non-sport activities that we excluded during the screening process. These reviews reported on yoga and mental health in clinical populations and not general populations or did not fit in our definition of organized activities, for instance because yoga was implemented as treatment (Balasubramaniam et al., 2012;Birdee et al., 2009;Cerrillo-Urbina et al., 2015;James-Palmer et al., 2020). Systematic reviews about group/choir singing and recreational rhythm music making in adults reported a positive impact on mental health outcomes (Noda et al., 2019;Williams et al., 2018;Yap et al., 2017). A recent umbrella review on performing arts, partly in the form of organized activities, in relation to many outcomes among adults also reported mixed findings regarding mental health outcomes (McCrary et al., 2021). This corresponds with the systematic reviews included in our umbrella review that yielded mixed findings on the impact of organized non-sport activities with mental (Cairns et al., 2014;Eime et al., 2013). The mixed findings could be because of a large variety in different organized non-sport activities. More systematic reviews and possibly more primary studies to examine the role of organized non-sport activities in relation to mental health in children and adolescents are needed due to the scarce available literature.
Most primary studies were performed in high-income countries. High-income countries have a large availability and variety in organized activities. Availability, affordability, transportation and cultural factors could affect participation (Mahoney et al.,). Contextual factors may influence the impact of organized activities on mental health in children and adolescents. In low and middle-income countries (LMICS) this may be different. Findings of this umbrella review may not be generalizable to LMICS. The included systematic reviews did not report differences regarding age or socioeconomic status. That no differences were reported does not imply that there are no differences. Caution is needed when generalizing these findings.
All but one systematic review received a critically low quality score by the AMSTAR-2. The AMSTAR-2 tool is stringent and rates systematic reviews as low quality in case of one critical weakness and as critically low in case of two critical weaknesses (Shea et al., 2017). Five systematic reviews have not explicitly mentioned an a priori protocol or deviated from their protocol and have not provided a detailed overview of excluded studies (with justification) at the stage of full text screening. Both critical weaknesses led to the critically low quality scores on the AMSTAR-2. If an a priori protocol or deviations were not mentioned this does not automatically mean that the study is biased, only that we do not know if it is. Providing an overview of excluded primary studies with justification shows readers that bias due to unjustified exclusion is not likely, and may increase the quality of systematic reviews. There is uncertainty regarding the bias of the systematic reviews but the results are consistent thus we feel confident about the findings. Confirmation of the reported results is needed. We emphasize the importance of adhering to the PRISMA guidelines (Page et al., 2020).
There is limited guidance on how to assess quality of evidence of outcome data in umbrella reviews (Gates et al., 2020;Pollock et al., 2016). Some umbrella reviews used the GRADE-tool (Grading of Recommendations, Assessment, Development, and Evaluation) (Pollock et al., 2016). This tool was developed for systematic reviews. For umbrella reviews the GRADE-tool could be applied by extracting the GRADE-assessment of primary studies (Pollock et al., 2016). If GRADEassessment was not applied in the systematic reviews it is not possible to use the GRADE-tool in the umbrella-review (Pollock et al., 2016). For our umbrella review we used a self-developed decision scheme to assess the quality of evidence. We call for further development of guidance on how to assess the quality of evidence when conducting an umbrella review. Our decision scheme in the appendix might serve as a first contribution.

Study limitations and strengths
This umbrella review has several strengths. We included systematic reviews about a variety of organized sport activities. We used a comprehensive search strategy and two researchers independently performed the selection process, data extraction and quality assessment. We included 118 primary studies with a low degree of overlap in the included systematic reviews. This indicates no unnecessary duplication over systematic reviews (Pieper et al., 2014). Rather that the included systematic reviews complemented each other. We carefully examined the quality of the included systematic reviews and were able to elicit important issues of published systematic reviews. We used a wide definition of mental health, as our objective was to aggregate all evidence about the impact of organized activities on mental health outcomes. We used a wide age range to ensure we did not miss any studies to children and adolescents. This umbrella review also has some limitations. For this umbrella review only systematic reviews with and without meta-analysis were considered eligible as these types of reviews provide the highest level of evidence. Consequently, this led to the exclusion of possibly relevant systematic searches or systematized qualitative, state-of-the art, narrative, mixed methods, overviews, rapid and scoping reviews (Grant and Booth, 2009). Findings of this umbrella review are dependent on the data and that has been synthesized and reported by the included systematic reviews. Some data were lacking in the included systematic reviews. It is unclear if this is due to data lacking in primary studies. Moreover, we have not found systematic reviews that studied a negative impact of organized activities on child or adolescent mental health. We have not restricted our search to positive outcomes of mental health. That we did not find any systematic review that studied the negative impact on mental health could be due to a focus on positive mental health outcomes in previous studies and to little focus on possible negative mental health outcomes. In the included systematic reviews most studies found a positive impact or null results whereas only a few reported a negative impact. However, earlier studies have postulated that too much participating in organized activities could lead to negative outcomes on mental health such as stress, particularly among affluent children and adolescents (Larson et al., 2006;Mahoney et al., 2006;Mahoney and Vest, 2012). This is also called the over-scheduling hypothesis (Larson et al., 2006;Mahoney et al., 2006;Mahoney and Vest, 2012). Possibly, there is bias in published research which could affect the appropriateness and interpretation of our findings.
This umbrella review identified a small positive impact of organized sport activities on mental health. Although making inferences based on the included systematic reviews is difficult due to their low quality of reporting on possible bias and different mental health outcomes, the findings are consistent over the systematic reviews included. Participating in organized sport activities can be stimulated by local policy. Thus, even though the magnitude of effect sizes is small, the impact may not be small if many children and adolescents will participate in organized sport activities. Approximately 40% of children and adolescents worldwide participate in sport (Tremblay et al., 2016). Preventive policies at the local level could contribute to better mental health by stimulating more children and adolescents to participate in organized sport activities. At municipal level, this can be done by for instance increasing the amount of local sport clubs and gyms. At school level, this can be done by for instance offering additional extracurricular activities (Somerset and Hoare, 2018). Policies could focus on increasing participation in organized sport activities using for instance social media campaigns or by collaborating with local sport organizations.(World Health Organization and Office, 2016).
Further research is needed to examine whether organized activities, particularly non-sport, contribute to better mental health. Research into the impact of organized non-sport activities on child and adolescent mental health is scarce and results are mixed. Based on the positive youth development theory a positive impact on mental health could be present (Fraser-Thomas et al., 2005;Lerner et al., 2005;Agans et al., 2016;Holt et al., 2020;Lerner et al., 2012). Future research may shed light on this possible association. More high-quality primary studies and more methodologically sound systematic reviews on organized nonsport activities may ensure this. Future research should disentangle the impact of participating in organized sport activities from mere physical activity. A focus on mental well-being is also warranted in future studies to determine if organized activities have a possible impact on mental health (Galderisi et al., 2015;Keyes, 2002Keyes, , 2005Keyes, , 2007Westerhof and Keyes, 2010;King et al., 2014;World Health Organization, 2004).

Conclusions
We found that there may be a small positive impact on mental health in children and adolescents participating in organized sport activities. This was not dependent on any specific type of organized sport activity. The observed findings should be interpreted cautiously in respect of the small effect sizes that were found, high heterogeneity of primary studies and possible publication bias. We cannot draw any conclusions about organized non-sport activities based on the small number of studies and the mixed results. Further research is needed to unravel possible mechanisms, possible mediation or confounding by physical activity and possible ways of implementing organized activities as positive preventive measure for child and adolescent mental health. This needs to be elaborated on particularly for organized non-sport activities.