How Can Physical Activity Be Promoted Among Children and Adolescents? A Systematic Review of Reviews Across Settings

1With recent studies showing that 124 million children worldwide are obese, physical activity (PA)—alongside nutrition—is a key in the fight against childhood obesity. Nevertheless, 75% of boys and 85% of girls in Europe do not fulfil the World Health Organisation (WHO) recommendations of at least 60 min of moderate-to-vigorous-intensity PA per day. As the impact of PA on children's physical, psychosocial, and intellectual development is well proven, PA promotion is a highly relevant topic.

2On the political level, central organisations are calling for action in order to combat childhood obesity. In 2011, the General Assembly of the United Nations declared that non-communicable diseases are one of the largest challenges of the twenty-first century, and that all sectors need to generate effective responses for the prevention and control of non-communicable diseases. Additionally, the European Union published an Action Plan on Childhood Obesity 2014–2020, and the WHO formulated specific recommendations in their report “Ending Childhood Obesity”. Furthermore, efforts specifically focusing on PA promotion are increasing. For example, the European Union Physical Activity Guidelines recommended a number of policy actions, while the WHO developed a European Physical Activity Strategy for 2016–2025.

3Due to the high relevance of this topic in politics and public health, a large number of reviews have been conducted to identify effective interventions for PA promotion. From the perspective of evidence-based medicine, the preferred method of knowledge summarisation is through the conduction of a systematic review of reviews. Such evidence is needed as the basis for decision making and allows research to have an impact on policy and practice.

4Over the past years, interventions for PA promotion have been investigated in reviews of reviews. Steenbock et al. analysed the efficacy of interventions to promote PA and healthy eating in the childcare setting, while Kriemler et al. dealt with the effects of school-based interventions on PA and fitness in children and adolescents. Moreover, the overarching review of Heath et al., which investigates many different types of PA interventions (not only for children and adolescents), also includes findings on school-based interventions.

5In contrast to the above reviews, our systematic review of reviews summarises the best available evidence for the target group of children and adolescents across the settings of family and home, childcare and school. Such information is required for the drafting of policy recommendations on the use of intervention strategies in settings to promote PA among children and adolescents. Another reason for obtaining such information is to identify potential “best buys” for PA promotion. Additionally, our review of reviews across settings was conducted to shed light on potential research gaps. The results were not only utilised to synthesise evidence, but have also formed parts of the German recommendations for PA promotion. To our knowledge, this review of reviews is the largest to be conducted with analysis on the efficacy and effectiveness of PA promoting interventions for children and adolescents.

6This systematic review of reviews was part of a large-scale project to develop national recommendations for PA and PA promotion. The overarching aim was to provide recommendations by target group: children and adolescents, adults, older adults, adults with a chronic disease, general population. This article provides an in-depth overview of results for the target group “children and adolescents.”

7In order to formulate recommendations that centre on the efficacy and effectiveness of interventions, a systematic review of reviews was conducted. Six electronic databases (PubMed, Scopus, Sport Discus, PsycInfo, ERIC, IBSS) were searched in 2015 for all publication years using the following search terms: “physical activity,” “intervention,” “evidence,” “effect,” “health,” and “review.” Alternative terms (e.g., bike, biking, cycling, walking, active transport, human powered transport, sedentary, exercise, sport) were defined and MESH-terms were formulated. Once the relevant literature had been identified, titles and abstracts were screened by two independent reviewers. The screening process was based on the following criteria: (a) The review contains empirical results from single studies. (b) The review includes interventions centred on the promotion of PA or the reduction of inactivity. (c) The review focuses on the efficacy and/or effectiveness of interventions. (d) Some of the single studies included in the review are of longitudinal design. (e) Reviews were written in English or German. Duplicates were excluded.

8The titles and abstracts of the identified records were screened to ensure that the above inclusion criteria were met. Two reviewers independently screened full texts from 223 reviews in a secondary screening process. Additional hand searches were conducted to identify further reviews. The remaining 213 reviews were then categorised by target group (children and adolescents, adults, older people, people with preconditions, general population).

9Seventy-four reviews dealt with the target group of children and adolescents (see Figure 1). Out of the 74 reviews, 51 were identified as a part of this systematic review of reviews, and 23 were obtained via hand search. Each of the 74 reviews underwent analysis. During the process of formulating national recommendations for PA, reviews were excluded due to the following reasons: reviews did not fulfil the inclusion criteria, did not deal with intervention studies, did not focus primarily on PA, were conducted unsystematically, only included a few studies dealing with PA promotion, were outdated or reported limited results, or were of very low quality. Once the exclusion process was complete, the total number of remaining reviews consisted of 39.

10One independent researcher assessed the quality of reviews using the AGREE instrument. This instrument was utilised in the formulation of the Canadian Physical Activity Guidelines.

11The AGREE instrument comprises of five criteria that deal with the methodological quality of studies. However, as the AGREE criteria are relatively easy to fulfil, methodologically weak reviews also receive high ratings. To counter this issue, an additional instrument was developed to enhance the accurate assessment of methodological quality for the identified reviews and meta-analyses (see Table 1). In order to assess review quality, both instruments were applied by calculating percentage values for each review. These values showed the percentages of fulfilled criteria per review, both on the basis of the AGREE criteria and the criteria from our newly developed instrument. The percentage values recorded for each review were calculated based on applicable criteria (e.g., some criteria were only applicable for meta-analyses).

12Combining the results of AGREE and our instrument ensured reliability when differentiating between the quality of reviews (especially for high quality reviews), thus improving the overall quality of assessment. Based on the combined results obtained from both instruments, the quality of each review was defined as high, medium or low.

13Following the methodology proposed by Smith et al., reviews were evaluated independently by two researchers. The evaluation focused on summarising evidence pertaining to various intervention types.

14A narrative synthesis of these 39 reviews is presented in this manuscript. The synthesis presents the results regarding the efficacy and effectiveness of PA promoting interventions. While efficacy trials test whether an intervention works under optimum conditions, effectiveness trials test an intervention under real-world conditions. However, even though the distinction of efficacy and effectiveness is highly relevant in health promotion research, both terms are often used interchangeably. When summarizing the main findings of the authors in our tables we overtook their wording, so that it might seem that many reviews analysed the effectiveness of interventions. In reality, and as stated in the discussion, most reviews analysed their efficacy.

15Table 2 provides an overview of the quality of reviews analysed in different settings.

16Three systematic reviews were identified that deal specifically with interventions in the family and home setting (Table 3).

17All three reviews stated that parental encouragement and support can increase PA in children, as well as reduce sedentary behaviour. Two reviews showed that parental behaviour influences the behaviour of their child. While one review identified a positive association between children's PA levels and the PA levels of their parents, another review found that a reduction in parental (sedentary) screen time can lead to decreased screen time in children. One review concluded that the level of parental involvement is more important than the setting in which the intervention is conducted.

18The availability of PA equipment seemed to have a positive effect on children's PA levels, whereas busy work schedules and parent fatigue demonstrated negative effects.

19Inconsistencies were found regarding the influence of parental enjoyment, maternal depressive symptoms, maternal self-efficacy, parental rules limiting sedentary behaviour, parental perceptions on the importance of PA, the level of child safety, and the child's physical competence.

20Four of the identified reviews dealt with PA promotion in childcare facilities (Table 4). In general, the majority of interventions for pre-schoolers seem to be centre-based and teacher-delivered.

21The reviews stated that theory or model-based interventions are more successful. The review by Ling et al. stressed that most theory or model-based interventions show a positive effect. In contrast, only one third of non-theory-based interventions were effective. The authors also indicated that the use of appropriate behavioural change strategies may be more important than theoretical models.

22Furthermore, based on the available evidence, Ling et al. concluded that multi-component interventions are more successful than single-component interventions. One of these components should be a structured PA programme. Interventions that had a positive effect on PA behaviour in children integrated structured PA into the formal curriculum. Both reviews stressed that the increase of structured PA should not be implemented at the expense of children's play time.

23Additionally, as children are most active during the first 10–15 min of being outdoors, experts suggest the provision of more free time. Another way to alter the free play environment at preschool is to provide additional play equipment—when play equipment isn't provided, children are more prone to sedentary behaviour and engagement in games that promote inactivity.

24To conduct interventions for PA promotion, Mehtälä et al. recommended PA-specific in-service teacher training. Both experience and personal characteristics played an important role in the promotion of PA among children in the childcare setting. Furthermore, Ward et al. stated that there is a need for teacher training, and that structured PA programs should be implemented by staff who are specialised and well-trained. Having well-qualified staff could also support skill development and competence in children.

25Two reviews recommended involving parents, while one review stated that evidence on parental support “remains unclear”. Nevertheless, Mehtälä et al. also stated that family is the most influential setting for young children, and that a partnership between families and childcare is crucial. Regarding specific strategies for parental involvement, providing health promotion information to parents is recommended in one review.

2628 of the identified reviews focused on the school setting (see Table 5). Strong evidence indicates that multi-component interventions are effective in the promotion of PA. While multi-component interventions are more effective than single-component interventions they do not seem to produce synergistic results.

27As one part of a multi-component approach, a range of reviews stated that interventions targeting physical education lessons are effective. Firstly, evidence shows that increasing the number of PE lessons is a key strategy to promote PA in school children. Secondly, improving the quality of PE lessons can be effective. Lonsdale et al. indicated that interventions can increase time spent in moderate-to-vigorous PA by 24% in PE lessons, thus substantially influencing the total amount of PA. Third, and strongly connected to higher quality PE lessons, reviews specified teacher training and capacity building as effective strategies to promote PA in school children.

28Furthermore, several reviews provide evidence for the integration of (more) PA into the curriculum. Evidence on integrating (PA-related) health education into the curriculum is inconclusive. While the integration of health education into the curriculum was recommended in one of the reviews, another review stated that results were mixed.

29Activity breaks—the integration of short bouts of PA into organisational routine—is another effective strategy to promote PA in the school setting, and has demonstrated “modest but consistent benefits”.

30After-school programs focused on PA and/or sports are also effective. In this context, improving students' attendance rates is highly important as the effectiveness of a program depends strongly on attendance.

31Changing the school environment also has an influence on PA in school children and is recommended in numerous reviews. In particular, the provision of equipment for games and playgrounds has proved effective. One review stated that the use of playground markings can increase PA in children during recess and lunchtime, whereas another study found inconclusive evidence. In one review, significant associations were found between decreased playground density and PA in children. Evidence on the effectiveness of interventions that only focus on environmental changes is limited; such interventions should be integrated in multicomponent approaches.

32Numerous reviews reported positive effects when involving parents and families in school-based interventions. Only one review indicated that evidence for parental involvement was inconclusive due to a lack of studies. One review of reviews mentioned that most but not all included records support the effectiveness of family involvement.

33Community involvement was linked with positive outcomes in two reviews. Surprisingly, one review stated that school-based interventions employing a community component were ineffective.

34The involvement of pupils is another relevant factor, as peer support can increase PA levels in school children. Adjusting interventions to this specific target population is necessary.

35Promoting active transport to school is another effective strategy to promote PA among children and adolescents. This strategy is related to promising policies such as ensuring safe routes to school and improving urban design.

36Finally, a number of reviews in the school setting addressed whether an intervention should acknowledge several behavioural components. While Pate et al didn't find any differences between interventions addressing single or multiple behavioural components, five other reviews did. The evidence shows that interventions addressing more than one health behaviour are less effective. The integration of other components in addition to PA was considered as a “stumbling block for success”. This point is stressed in another review that focuses on after-school programs. In general, intervention effectiveness is higher when focus is placed on a specific goal compared to interventions that require broader focus.

37Four reviews were classified as “other” as their results were not connected with one specific setting (Table 6).

38Three of these reviews analysed a broad range of health promoting interventions. As the reviews dealt with interventions not only pertaining to PA, results on efficacy are not as specific as other reviews. Overall, the reviews stated that health promotion interventions can be effective in terms of health, obesity prevention, and the reduction of socioeconomic inequalities in obesity. One of the reviews stated that interventions in the healthcare setting can be effective.

39One of the reviews analysed computer- and web-based interventions in both the home and school setting. This intervention type can be effective, especially with regard to schools.

40This systematic review of reviews provided important findings regarding PA promotion for children and adolescents: Parents play a key role in the family and home setting. In settings such as childcare facilities and schools, evidence shows that parental involvement is an important factor for the efficacy of interventions.Multi-component interventions proved to be effective in the childcare setting and the school setting.For childcare facilities, reviews stated that interventions should be theory or model-based, include a PA programme, provide more free time, and contain teacher training.In schools, evidence is available for increasing the number of PE lessons, as well as improving the quality of PE lessons, teacher training and capacity building. Furthermore, findings indicate the efficacy of other intervention strategies: implementation of more PA into the curriculum; activity breaks; after-school programs; changes in the school environment; promotion of active transport; community and peer involvement.

41Parents play a key role in the family and home setting. In settings such as childcare facilities and schools, evidence shows that parental involvement is an important factor for the efficacy of interventions.

42Multi-component interventions proved to be effective in the childcare setting and the school setting.

43For childcare facilities, reviews stated that interventions should be theory or model-based, include a PA programme, provide more free time, and contain teacher training.

44In schools, evidence is available for increasing the number of PE lessons, as well as improving the quality of PE lessons, teacher training and capacity building. Furthermore, findings indicate the efficacy of other intervention strategies: implementation of more PA into the curriculum; activity breaks; after-school programs; changes in the school environment; promotion of active transport; community and peer involvement.

45Additionally, several gaps were identified in the evidence base. These gaps partly reflect underlying controversies regarding research paradigms utilised to study effects of PA interventions: Research focuses primarily on individual level interventions even though community and policy level interventions are considered to be more effective in terms of public health. Such focus on individual level interventions can be explained by the predominance of evidence-based medicine for the evaluation of interventions. Within the paradigm of evidence-based medicine, it is of high importance to generate knowledge through randomised and controlled trials. These study designs, however, have low applicability if one wants to examine policy or environmental interventions. Due to this, a combined use of systematic reviews, meta-analysis and realist synthesis as employed in a recent analysis of family-based PA interventions seems to be beneficial.Research is dominated by studies investigating interventions in the school setting. Interventions in the settings of family and childcare were only investigated in a limited number of reviews—even though there is a lack of knowledge regarding interventions outside of the school setting. One review couldn't identify a single study including effective interventions for children aged 0–5 years. In particular, for nations where children only spend half a day in school (for example, in Germany), school-based interventions might have less relevance for PA promotion. This is because schools have neither the capacities nor the facilities for additional interventions for PA promotion. For such nations, more knowledge is needed on PA promotion outside of the school setting.Research on interventions in settings other than family, childcare and school is almost non-existent. As other studies show that computer- and web-based interventions can be effective in promoting PA of children and adolescents, this gap in the evidence base could limit public health impact. Moreover, we could not find any evidence regarding the sport club setting. In particular, for nations which have a sport club-based system for PA in leisure-time, such knowledge could prove valuable.Most reviews investigated the “efficacy” of interventions with very limited information on “effectiveness”. This is problematic as researchers raise the question whether interventions that are successful in efficacy studies are also effective in the real world. From a Public Health perspective, study designs that allow for the simultaneous testing of both efficacy and effectiveness—such as pragmatic trials—might be more appropriate to generate this evidence. This could allow accelerating scale-up processes of interventions. Also the RE-AIM framework can be used for analysing both efficacy and effectiveness of interventions, as recent systematic reviews of physical activity interventions for children and youth show.

46Research focuses primarily on individual level interventions even though community and policy level interventions are considered to be more effective in terms of public health. Such focus on individual level interventions can be explained by the predominance of evidence-based medicine for the evaluation of interventions. Within the paradigm of evidence-based medicine, it is of high importance to generate knowledge through randomised and controlled trials. These study designs, however, have low applicability if one wants to examine policy or environmental interventions. Due to this, a combined use of systematic reviews, meta-analysis and realist synthesis as employed in a recent analysis of family-based PA interventions seems to be beneficial.

47Research is dominated by studies investigating interventions in the school setting. Interventions in the settings of family and childcare were only investigated in a limited number of reviews—even though there is a lack of knowledge regarding interventions outside of the school setting. One review couldn't identify a single study including effective interventions for children aged 0–5 years. In particular, for nations where children only spend half a day in school (for example, in Germany), school-based interventions might have less relevance for PA promotion. This is because schools have neither the capacities nor the facilities for additional interventions for PA promotion. For such nations, more knowledge is needed on PA promotion outside of the school setting.

48Research on interventions in settings other than family, childcare and school is almost non-existent. As other studies show that computer- and web-based interventions can be effective in promoting PA of children and adolescents, this gap in the evidence base could limit public health impact. Moreover, we could not find any evidence regarding the sport club setting. In particular, for nations which have a sport club-based system for PA in leisure-time, such knowledge could prove valuable.

49Most reviews investigated the “efficacy” of interventions with very limited information on “effectiveness”. This is problematic as researchers raise the question whether interventions that are successful in efficacy studies are also effective in the real world. From a Public Health perspective, study designs that allow for the simultaneous testing of both efficacy and effectiveness—such as pragmatic trials—might be more appropriate to generate this evidence. This could allow accelerating scale-up processes of interventions. Also the RE-AIM framework can be used for analysing both efficacy and effectiveness of interventions, as recent systematic reviews of physical activity interventions for children and youth show.

50From an international perspective, the key findings and gaps in the evidence base are associated with several challenges for PA promotion. Such challenges are mainly caused due to focus being placed on the school setting: Firstly, the concrete implementation of effective interventions depends on the structure of the educational system. In a study comparing the educational governance of the USA and France, the authors stated that “institutional sectors in liberal polities are often organised as complex multi-layered governance systems characterised by fragmented decision-making structures.” In state-centred polity on the other hand, “one can expect more tightly structured institutional sectors”. These findings affect the implementation of the results from this systematic review of reviews into practice (e.g., with regard to the number and quality of PE lessons). In Germany, the federal system's fragmented decision-making structure allows federal states to decide on school curricula. In Hungary however, regulation on daily PE classes was implemented as a nation-wide policy in 2012/2013.Secondly, interventions that are most needed might also be connected with overall school policy. In Germany, school lessons usually finish at 1 p.m. or 2 p.m. However, the percentage of pupils visiting all-day schools has constantly risen over the years (from 9.8% in 2002 up to 39.3% in 2015). With an increasing number of pupils visiting all-day schools, the need for additional sport facilities, and changes in the school environment might be most relevant.For school children, the integration of sectors other than education is only investigated as part of a multi-component approach. The necessity of focusing on the family setting is stressed by findings on the importance of parental involvement. For the promotion of active transport to school, the influence of urban planning requires further investigation.The question of how to finance interventions in the education sector was not even raised in the reviews, even though the lack of public funding and resources is perceived as a barrier for childhood obesity prevention by two thirds of stakeholders in Europe. Intersectoral partnerships might be a promising approach to finance programs with combined resources.

51Firstly, the concrete implementation of effective interventions depends on the structure of the educational system. In a study comparing the educational governance of the USA and France, the authors stated that “institutional sectors in liberal polities are often organised as complex multi-layered governance systems characterised by fragmented decision-making structures.” In state-centred polity on the other hand, “one can expect more tightly structured institutional sectors”. These findings affect the implementation of the results from this systematic review of reviews into practice (e.g., with regard to the number and quality of PE lessons). In Germany, the federal system's fragmented decision-making structure allows federal states to decide on school curricula. In Hungary however, regulation on daily PE classes was implemented as a nation-wide policy in 2012/2013.

52Secondly, interventions that are most needed might also be connected with overall school policy. In Germany, school lessons usually finish at 1 p.m. or 2 p.m. However, the percentage of pupils visiting all-day schools has constantly risen over the years (from 9.8% in 2002 up to 39.3% in 2015). With an increasing number of pupils visiting all-day schools, the need for additional sport facilities, and changes in the school environment might be most relevant.

53For school children, the integration of sectors other than education is only investigated as part of a multi-component approach. The necessity of focusing on the family setting is stressed by findings on the importance of parental involvement. For the promotion of active transport to school, the influence of urban planning requires further investigation.

54The question of how to finance interventions in the education sector was not even raised in the reviews, even though the lack of public funding and resources is perceived as a barrier for childhood obesity prevention by two thirds of stakeholders in Europe. Intersectoral partnerships might be a promising approach to finance programs with combined resources.

55Based on this comprehensive and up-to-date cross-sectoral review, national recommendations for physical activity promotion were developed in Germany. These recommendations aim to have an impact on public health by offering scientific orientation for experts and stakeholders. Alongside the efficacy of interventions, the national recommendations also consider aspects regarding PA promotion in children and adolescents: Effectiveness: In terms of public health, did the intervention prove to be effective on a large scale?Health equity: Is the intervention able to address and reduce health inequalities?Cost effectiveness: Does the intervention demonstrate a good relation between costs and the expected benefits?Quality criteria: Which criteria need to be considered to ensure the successful implementation of an intervention?

56Effectiveness: In terms of public health, did the intervention prove to be effective on a large scale?

57Health equity: Is the intervention able to address and reduce health inequalities?

58Cost effectiveness: Does the intervention demonstrate a good relation between costs and the expected benefits?

59Quality criteria: Which criteria need to be considered to ensure the successful implementation of an intervention?

60Considering these aspects is highly beneficial for bridging the gap between evidence and practice. For decision makers, effectiveness and health equity are important criteria, as well as cost effectiveness. Furthermore, quality criteria are important for practitioners, and decision makers: For example, evidence shows that the involvement of all relevant stakeholders—children, families, teaching staff and management—in the planning of an intervention increases the likelihood of its success.

61In order to improve the evidence base, future research on PA promotion for children and adolescents should focus on the above-mentioned aspects.

62AR acquired the funding, administrated the project, developed concept and methodology of this study and analysed the data. SM supported both literature search and data analysis and wrote the original manuscript draft. KA-O oversaw the literature search and data analysis and supported the development of the original draft. UU-R performed data analyses. LG contributed to manuscript preparation through the revision of numerous drafts. IB supported the literature search. GG developed the quality assessment tool and assessed the quality of reviews. All authors critically reviewed and edited the manuscript.

63The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.