Full text

Turn on search term navigation

Introduction

Globally, about 18% (over 34 million) children and adolescents are overweight or obese–a 10-fold increase from 40 years ago [1]. Physical inactivity has been identified as one of the main risk factors whereby two-thirds of children and adolescents are insufficiently active, despite the widely recognized benefits of physical activity [2]. Evidence shows that inactive children are likely to become inactive adults [3, 4], and it is projected that the healthcare burden of physical inactivity-related non-communicable diseases will cost INT$520 billion annually between 2020–2030 if the physical inactivity pandemic continues [5].

Over the past three decades, the number of physical activity behaviour (PAB) and sedentary behaviour (SB) interventions targeting childhood inactivity has seen an upward surge [6]. PA is defined as any movement produced by skeletal muscles that involve the energy expenditure of >1.5 metabolic equivalents of tasks (METs) whereas ≤1.5 METs while awake is considered as SB [7]. Conclusions about the effectiveness of interventions for school-aged children from recent systematic reviews have been mixed [8–12]. Typically, these interventions aim to manipulate factors associated with PAB and/or SB, hence these factors are also considered as determinants as their causal associations with PAB or SB are assumed [13]. Not only should determinants be evidence-based, but they should also be modifiable to the extent that can enact behaviour change [14]. An array of determinants relevant to the youth population within the European context has been previously identified by experts of PAB and SB [14, 15]. Based on the socio-ecological model [16], the majority of the identified PAB/SB determinants (approximately 55%) considered to be highly modifiable and have the largest effect on PAB/SB, belong to the individual and interpersonal level, such as attitude, support of peers/family and TV exposure [14, 15]. However, there have been mixed findings on the extent to which interventions that target these determinants are associated with changes in PAB/SB [17–19]. A lack of understanding of which determinants have significantly contributed to changes in PAB/SB has hampered progress in physical activity promotion across the lifespan [20]. Given this state of uncertainty, the DEterminants of Physical Activity in SettingS (DE-PASS) consortium was formed with an aim to identify key determinants effective in promoting PAB and reducing SB, and crucially, translatable at the policy level to accelerate research-policy collaborations in addressing the physical inactivity pandemic.

Several factors may have contributed to the mixed findings regarding the association between the modifiable determinants and PAB/SB. First and foremost, the context within a setting in which the determinants operate is seldom considered [21]. Interventions are often complex for many reasons, including but not limited to the stakeholders involved and their motivation, the physical and psychological capacity for (long-term) implementation and the prevailing PAB/SB practice where the interventions are implemented. As such, the extent to which these factors may influence the modifiable determinants may vary considerably in different settings [22–24]. For example, results of realist reviews of interventions for children showed that in the family setting, physical activity knowledge combined with parental reinforcement was an important determinant unique to that setting, whereas parental restrictions on PAB, as a determinant, hampered the effect of school-based interventions [21, 22]. Another factor that warrants attention is the age groups included in reviews targeting youth, where interventions involving children and adolescents were examined collectively [10, 12, 14]. The developmental journey from childhood to adolescence sees notable changes and adaptations in individuals’ environmental, physical and psychological conditions, all of which define the individuals’ context [25]. For example, while self-efficacy was found to be a common modifiable PAB determinant for children and adolescents, intention appears to be unique to children and perceived behaviour control and planning are unique to adolescents [26, 27]. Given the above considerations, the current systematic review will examine PAB/SB determinants in interventions from different settings, targeting children aged 5–12 years only.

To address a main objective of DE-PASS of generating a Best Evidence STatement (BESt) with regards to the key modifiable determinants for youth PAB/SB from existing best evidence, the current review aimed to examine interventions that target PAB and/or SB using the randomized controlled trial (RCT) and controlled trial (CT) designs. While RCTs are considered the gold standard in intervention design, CTs could be a viable alternative when randomization is challenging due to factors such as participants’ or stakeholders’ preference. This review also focused on self-report PAB/SB measures only, while other planned systematic reviews addressing the same DE-PASS objective will focus on device-based measures, as the discrepancy in measurement is evidenced [28–30]. Therefore, the aims of this systematic review were three-fold–i) to identify the modifiable determinants that have been targeted in PAB/SB interventions in different settings, ii) to evaluate the extent to which these determinants have been modified, and iii) to investigate their association with self-report PAB/SB in school-aged children.

Methods

This review is one of the five planned systematic reviews conducted under the same deliverable (youth focus) within the DE-PASS consortium. Workshops for all members involved in the review activities were conducted to ensure mutual understanding of the eligibility criteria and the practice in study screening, data extraction, risk of bias assessments, and the use of Covidence, an online systematic review platform (www.covidence.org).

Study design

This systematic review was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) guidelines (S1 Checklist) [31]. The study protocol was prospectively registered in PROSPERO (CRD42021282874).

Search strategy

The current study applied the same search strategy for all the five systematic reviews under the same deliverable (youth focus) within DE-PASS. A search was conducted on MEDLINE, PsycINFO, Web of Science, Sport Discus and Cochrane Central Register of Controlled Trials for literature from 2010 up to July, 2023. We considered publications from 2010 because this was when WHO published the first global PA guidelines [32]. For the full search strategies and terms, please refer to the published study protocol http://dx.doi.org/10.1136/bmjopen-2021-059202 [33].

Eligibility criteria

Population.

We included children aged 5–12 years (inclusive) without known medical conditions that would hinder habitual PAB, such as spina bifida and arthritis.

Interventions.

Interventions targeting PAB/SB in children using self-report, and that i) had measured modifiable determinants at ≥2 time points (pre-/post-measurements), and ii) had measured the PAB/SB outcomes at ≥2 time points (pre-/post-measurements) were included.

Comparator.

All studies included a control group receiving no intervention, or a comparator group receiving an alternative intervention matched to the experimental conditions.

Outcomes.

This review included two types of outcomes–modifiable determinants and self-report PAB/SB, as the intervention effect on both were examined separately. We assessed whether an outcome is qualified as a determinant by the theoretical underpinning or the context of the interventions. For example, if an intervention explicitly aimed to reduce body weight in order to promote PAB, body weight status was considered a determinant. If, however, body weight was clearly considered as an outcome without specifying its mechanistic influence on PAB in the context of the intervention, and given no other modifiable determinants were included, the study was excluded. For studies with both self-report and device-based PAB/SB measurements, only the former was analyzed in this review.

Study design.

Interventions that followed RCT or CT designs of any duration and follow-up period, and within any settings, were analyzed. Peer-reviewed studies in any language were considered. For studies that did not provide relevant information for eligibility assessment or for data extraction, authors were contacted. These studies were excluded if the requested information was not obtained.

Study selection and data extraction.

At the initial screening, Endnote x9 was used to remove duplicates and non-peer-reviewed literature. The final identified studies were transferred to Covidence for title/abstract/full-text screening and data extraction. Extracted data included sample characteristics, study characteristics, settings, theoretical basis of the interventions, measurements of PAB/SB and determinants as well as their measurement properties. Study screening and data extraction were completed by deliverable members of DE-PASS in pairs independently. Conflicts were solved by discussion or with a third member.

Quality assessment

With regards to risk of bias assessments, we used the Cochrane Risk of Bias Tool for Randomized Trials version 2 (RoB2.0) [34] and Risk of Bias in Non-Randomized Studies of Intervention (ROBINS-I) [35] for RCTs and CTs respectively. The ‘Bias in the measurement of outcome’ domain was assessed for both outcomes of interest separately, namely determinant(s) and PAB/SB. Two independent reviewers assessed the risk of bias. A third reviewer was consulted if consensus could not be reached. The assessment plots were generated by the robvis tool [36]. To assess the certainty of evidence, three authors (FCML, AM, KN) followed the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to evaluate all studies included in the meta-analyses [37]. GRADE includes five criteria–risk of bias, inconsistency, indirectness, imprecision and publication bias. The level of certainty ranges from high to very low, depending on the extent to which the true effect is considered similar to the estimated effect.

Statistical analysis

All determinants were categorised based on the socio-ecological model [16]. For determinants with multiple indicators (e.g., different parental practices to minimize screen time), either the composite score was calculated (see S1 File) [38], or a total score provided by the authors was used. Conceptually similar determinants were grouped for analysis where possible.

For studies that reported multiple PAB/SB outcomes, the one that most reflected total daily PAB/SB was used (e.g., habitual MVPA). Composite scores of SB were calculated for SB outcomes that could be combined to reflect habitual SB (e.g., total daily screen time and total daily computer use).

To account for the possible co-variance of individual scores within each composite score, sensitivity analyses were conducted where different correlation coefficients were applied to the formula by Borenstein (2011), to test if effect size might change substantially [38]. Where results at multiple time points were recorded during an intervention, only results at post-intervention were considered (pre-post effect). Additionally, if more than one time-point follow-ups were reported, e.g., after three weeks, one month, and three months, the latest time-point results were used to reflect a longer-term effect (pre-follow-up effect).

Standard mean difference and standard error for changes in determinants and PAB/SB outcomes (from baseline to follow up) were calculated for each included study where possible (see S1 File for details). Individual studies were first inspected for corresponding determinant and PAB/SB changes. For determinants and PAB/SB that could be pooled for meta-analysis, Robust Bayesian meta-analysis (RoBMA) was conducted in JASP 0.16.4 [39, 40], which uses the RoBMA R package [41] and Markov Chain Monte Carlo algorithms via JAGS [42]. We used only random-effects part of the RoBMA model ensemble with the default prior distributions, resulting in 18 included models (detailed RoBMA specification can be found in [43]). We used Bayes factor (BF₀₁) to measure evidence of the absence of an effect over the presence of an effect. The same criteria were also applied to publication bias assessment. While the Bayes factor is a continuous measure of strength of evidence, we used the following rule of thumb to aid interpretation: 1<BF₀₁<3 = weak evidence (i.e., presence or absence of an effect cannot be ascertained), 3<BF₀₁<10 = moderate evidence, BF₀₁>10 = strong evidence for the null [44]. When evidence for the alternative was considered, the Bayes factor was simply inverted (e.g., BF₀₁ = → BF₁₀ = 3, which implies weak evidence for an effect). Cohen’s d with 95% credible interval (CI) was also reported. Cohen’s d ≥ 0.2 (small effect), ≥ 0.5 (moderate effect), ≥ 0.8 (strong effect) [45]. The degree of heterogeneity was assessed by the between-study standard deviation τ. Studies that could not be included in the meta-analyses were reported narratively. For readers unfamiliar with RoBMA, we supplemented the results of the corresponding indicators, including effect size (95% CI), heterogeneity and publication bias, using classical frequentist analysis with random effects meta-analysis and Vevea and Hedges (1995) selection model for publication bias correction (see S2 File) [46]. All intervention settings and outcomes (PAB vs SB, and habitual vs non-habitual PAB were analysed separately. RCTs vs CTs were also examined separately for the purpose of GRADE. The effect of interventions on determinants was analysed regardless of their PAB or SB outcomes.

Results

Study selection

After removing duplicates, 27,581 studies were subject to title and abstract screening. Screening resulted in 1,762 full texts to be assessed for eligibility. Out of the 184 eligible studies, 15 were deemed relevant to the current review (self-report PAB/SB measurement with children 5-12yrs) (Fig 1). Excluded studies can be found in S1 Data, and full extracted data with data extractor information and data extraction date can be found in S2 Data.

[Figure omitted. See PDF.]

Study characteristics

Table 1 shows the study characteristics and participant demographics of the 15 studies included in this review (10 RCTs, five CTs), totalling 13,107 participants. The settings were determined based on where the interventions were delivered. Four settings were identified from this pool–school only, family only, school with family/home and community with/without other settings (e.g., family/home, school). Interventions that took place in the home environment but without involving parents explicitly, and those that targeted parental involvement, were all classified as family/home setting. Thirty-seven distinct determinants were identified– 25 were individual (psychological) determinants, seven were interpersonal (psychological) determinants, two were individual (behavioural) determinants, another two were interpersonal (behavioural) determinants and one was institutional determinant. Further study characteristics and evidence synthesis for each setting are provided in Table 1. Effect size and 95% CI for each determinant and PAB/SB outcome by settings can be found in S1 Appendix).

[Figure omitted. See PDF.]

(I) School setting

Five studies–one RCT and four CTs–with sample sizes ranging from 179 to 1,464 were identified. Intervention duration ranged from eight weeks to two years, and follow-up periods from the end of the interventions ranged from eight weeks to 15 months. Four interventions targeted changes in PE curricula and/or PE teacher training [47–50], and one intervention implemented a video game-based program at schools [48]. Three interventions were theory informed [48, 50, 51]. Four studies measured habitual PA [47–49, 51], and one study measured leisure time PA [50], using validated instruments (see Table 1).

Study outcomes—PAB/SB and determinants.

RCT. Six conceptually different determinants were targeted in the RCT, of which four belonged to the individual (psychological) category and two belonged to the interpersonal (psychological category) [50]. The psychometric properties of all determinant measurements were referenced (Table 1). There was no significant change in determinants at immediate post-intervention and at follow-up (d’s ranged from -0.17 to 0.28), except a small effect on students’ perception of teachers’ controlling behaviour at follow-up (d = 0.25, 95%CI 0.12 to 0.37), indicating teacher’s behaviour was perceived to be more controlling which is against what the intervention aimed to achieve. Additionally, there was non-significant effect on PAB (d = -0.02).

CTs. Five distinct individual (psychological) determinants were targeted in all four studies. The psychometric properties of all determinant measurements were referenced (see Table 1). Only self-efficacy from one study showed a small effect at post-intervention (d = 0.40; 95% CI 0.09 to 0.73) [51], however, when it was pooled in a meta-analysis with another study [44], there was moderate evidence against an effect on self-efficacy (Table 2; Fig 2A). Narratively, there was moderate intervention effect on commitment to PA (d = 0.68; 95% CI 0.50 to 0.87) [49]. However, none of the determinants that could only be analysed narratively reported notable intervention effect (d’s ranged from -.17 to 0.28; for determinants that showed small effects, the 95% CI’s crossed the estimate threshold).

[Figure omitted. See PDF.]

Forest plots depicting intervention effect on (a) self-efficacy, (b) overall PAB pre-/post- effect and (c) overall PAB pre-/follow-up effect in CTs under school setting.

[Figure omitted. See PDF.]

The effect size estimates for meta-analyses and heterogeneity are expressed in d (95%CI) and τ respectively.

For the individual study that showed a small effect on self-efficacy at post intervention, there was a corresponding small effect on PAB at the same time point [51]. For the study that measured commitment to PA, there was no corresponding effect on PAB. When all CTs were pooled for meta-analysis, there was moderate evidence against an effect on PAB at post-intervention (Fig 2B), however, there was insufficient evidence to suggest presence or absence of an effect on PAB at follow-up, or publication bias (Table 2; Fig 2C). Heterogeneity for all meta-analyses seem small, but due to limited number of studies in each meta-analysis, the degree of heterogeneity is highly uncertain.

Quality assessment

For the four CTs, they were all deemed high risk of bias overall. Notable contributors to the judgement were three domains—domain 1 (bias due to confounding), domains 6 and 7 (bias due to measurement of outcomes–PAB/SB and determinants respectively). Judgement for domain 1 primarily stemmed from the fact that not all pre-defined confounders were accounted for in all studies, while judgement for domains 6 and 7 was because participants were unlikely to be blinded in most interventions involving self-report measurements (PAB/SB and determinants) (Fig 3A). Nonetheless, the one RCT in this setting explicitly mentioned blinding of the participants and researchers (Fig 3B) [50].

[Figure omitted. See PDF.]

Risk of bias assessments of CTs using (a) Robins-I and risk of bias assessments of the RCT using (b) RoB2.0 in school setting.

Certainty of evidence and intervention effect.

We conducted GRADE for the meta-analysis of self-efficacy in two CTs (Table 3A) [47, 51], and the meta-analysis with PAB as an outcome in four CTs at post-intervention [47–49, 51]and the two CTs at follow-up [49, 51] (Table 3B). The certainty of evidence was high for the absence of intervention effect on PAB at post-intervention, whereas for the other two meta-analyses, the certainty of evidence was deemed low mainly due to imprecision of effect estimate.

[Figure omitted. See PDF.]

b. Overview of quality of evidence (GRADE) and intervention effect on physical activity for 4 CTs in school setting.

(II) Family/home setting

Only one intervention was conducted in the family/home setting (see Table 1) [52]. It was a theory-based 24-week RCT with no follow-up assessment, and both habitual PAB and SB were examined (n = 251). Three determinants were measured using validated instruments–two belonged to individual (psychological) and one belonged to interpersonal (behavioural) categories. No significant intervention effects were reported for all outcomes (d’s ranged from -0.17 to 0.10). The risk of bias was deemed high due to bias in the measurement of outcomes (domain 4 and 5 –PAB/SB and determinants respectively) (Fig 4).

[Figure omitted. See PDF.]

(III) School with family/home settings

We identified seven studies of which one was a CT [53], and all interventions were theory-informed (see Table 1). Sample sizes ranged from 51 to 3,147. Intervention period ranged from six weeks to 20 months, and only 1 study included follow-up (10 weeks post-intervention) [53]. Four interventions indirectly involved parents in the form of homework completion with children or remote knowledge provision [54–57], two interventions actively involved parents in workshops or information sessions [53, 58], and one intervention relied on children adhering to the home intervention [59]. Habitual PAB [53, 55, 56] and habitual MVPA [53] were measured in the respective studies, and SB was measured in four studies [54, 57–59].

Study outcomes—PAB/SB and determinants.

RCTs. Twenty-three conceptually different determinants were targeted in these interventions (see Table 1). Fifteen determinants belonged to the individual (psychological) category, one belongs to individual (behavioural) category, five belonged to the interpersonal (psychological) category, one belongs to the interpersonal (behavioural) category and one belonged to the institutional category.

Out of all RCTs, determinants that showed positive effects were–i) parental support (d = 0.24, 95%CI (0.09, 0.40)) [52],external motivation (d = -0.23, 95%CI -0.38 to -0.07) [55] and parental modelling on SB (d = 0.25, 95%CI 0.18 to 0.32) [57]–small effect; ii) knowledge (d = 0.50, 95%CI 0.31 to 0.68) [58]–moderate effect, and iii) self-efficacy (d = 0.90, 95%CI (0.31, 1.57)) and exercise intention (d = 0.87, 95%CI 0.30 to 1.45) [56]–strong effect (Table 1). Other determinants showed non-significant intervention effect (d’s ranged from -0.44 to 0.52; for determinants that showed small to moderate effects, the 95% CI’s crossed the estimate threshold) [54–59].

Five determinants were targeted in more than one study–self-efficacy [56, 57, 59], attitude [56, 57], subjective norm [56, 57], knowledge [56, 58] and parental practice in SB regulation [54, 57], hence we conducted a meta-analysis for each (Fig 5A–5E). Results of meta-analyses showed moderate evidence against an effect on self-efficacy, attitude, subjective norm and parental practice (Table 4). There is moderate evidence of presence of publication bias for the meta-analysis on self-efficacy (BF₁₀ = 5.88). Evidence for presence or absence of an effect on knowledge and publication bias for other meta-analyses cannot be determined.

[Figure omitted. See PDF.]

Intervention effects on (a) self-efficacy in studies targeting PAB and/or SB; (b) attitude in studies targeting PAB and/or SB; (c) subjective norm in studies targeting PAB and/or SB; (d) knowledge in studies targeting SB; (e) parental practice in SB regulation in studies targeting SB; (f) PAB at post-intervention; (g) SB at post-intervention, in RCTs under school and family/home settings.

[Figure omitted. See PDF.]

The effect size estimates for meta-analyses and heterogeneity are expressed in d (95%CI) and τ respectively.

For intervention effect on PAB/SB, meta-analyses showed moderate evidence against an effect on PAB (Fig 5F) and strong evidence against an effect on SB (Fig 5G). There was moderate evidence for publication bias for PAB (BF₁₀ = 7.69). In one study included in the meta-analysis for PAB and for knowledge [58], moderate effect on knowledge (the only determinant measured) and small effect on PAB (d = 0.30, 95%CI 0.12 to 0.48) [58] was found, however, the psychometrics of knowledge measurement was not referenced. Therefore, despite that some interventions showed promise on narratively analysed determinants (i.e., external motivation [55], exercise intention [56], and parental modelling [57]), corresponding change on the pooled PAB/SB effect is not evident. Again, the magnitude of heterogeneity of all meta-analyses appeared small, but this remains inconclusive due to small number of studies in each meta-analysis (Table 4). Together with other results that did not see any corresponding changes in determinants and PAB/SB, we could only suggest that associations between these determinants and PAB/SB were possible, and that the interventions had not been successful in changing either.

CT. The only CT under this setting showed small effect on perceived barriers to PA (d = 0.43, 95%CI 0.06 to 0.80) at post-intervention and self-management (d = 0.43, 95%CI (0.06, 0.81)) at follow-up. There was a moderate effect on social support from schools at follow-up (d = 0.53, 95%CI 0.16 to 0.91). However, PAB notably decreased at post-intervention (d = -0.51, 95%CI -0.88 to -0.13) and there was negligible intervention effect on PAB at follow-up (d = 0.00) [53].

Quality assessment

All studies are deemed high risk of bias, mainly due to bias in the measurement of outcomes (PAB/SB and determinants) (Fig 6).

[Figure omitted. See PDF.]

Risk of bias assessments of (a) RCTs and (b) CT in school with family/home settings.

Certainty of evidence and intervention effect.

We conducted GRADE for the seven meta-analyses on self-efficacy, attitude, subjective norm, knowledge, parental practice, PAB and SB (all RCTs; Table 5A and 5B). The certainty of evidence was downgraded to low or very low for all meta-analyses due to risk of bias and/or imprecision.

[Figure omitted. See PDF.]

b. Overview of quality of evidence (GRADE) and intervention effect on physical activity and sedentary behaviour in school with family/home setting.

(IV) Community with/without other settings

Two RCTs were identified (see Table 1). One intervention was conducted in a community setting only with an active control group [60], and one was in the community with family/home and school setting (where parents were indirectly involved in the intervention and active policy/environmental adaptation was in place) [61]. Sample sizes ranged from 71 to 1,348. Interventions lasted from four weeks to two years, and follow-up periods ranged from three to six months. Only one intervention was theory-informed [60]. One study measured habitual PAB and SB [60] and one targeted transportation PA [61], only the former measurement instruments was referenced.

Study outcomes—determinants and PAB/SB.

Within the two studies, six determinants were targeted–five were individual (psychological) and the other one was interpersonal (psychological). The measurements, and measurement properties, of three determinants within one study were unspecific [56] (see Table 1).

Due to difference in the type of PAB/SB measured [60, 62], PAB/SB outcomes can only be analysed descriptively. The community-based intervention showed a moderate post-intervention effect on SB (d = -0.74, 95%CI -1.22 to -0.26), with trivial effects on all determinants (d’s ranged from 0.18–0.33; for determinants that showed small effects, the 95%CI’s crossed the threshold). However, the small effect on MVPA/SB expectations at follow-up (d = 0.39, 95%CI 0.04 to 0.75) did not see any changes to PAB or SB [60]. For the other study, there were only trivial effects for perceived safe route to school, parental encouragement for cycling to school, attitude towards cycling and % active trips to school (d’s ranged from -0.14 to 0.14) [61].

Quality assessment

Similar to other studies included in this review, the main reason for the overall high-risk decision on study bias was due to measurement bias [61]. However, the study with an active control group could contribute to blinding of participants, as such, the study was deemed low risk (Fig 7).

[Figure omitted. See PDF.]

Sensitivity analyses

Sensitivity analyses were performed for studies that required estimations of composite scores. No difference in the effect sizes was detected when r was set at 0.2, 0.5 and 0.8.

Discussion

The main aims of this review were to identify modifiable determinants that have been targeted in interventions that followed the RCT/CT designs and to assess their association with self-report PAB/SB in children in different settings. To our knowledge, it is the first study to apply Robust Bayesian meta-analyses to examine the effects of interventions on modifiable determinants, and to infer the associations between the determinants and PAB/SB where possible. Out of the 37 distinct determinants targeted across all settings, 68% were individual (psychological) determinants, 5% were individual (behavioural) determinants, 20% were interpersonal (psychological) determinants, 5% were interpersonal (behavioural) determinants and there was only one institutional determinant. Common determinants across settings were self-efficacy, family support, school support, peer support, motivation based on self-determination theory, PA enjoyment, caregivers’ PA, perceived competence and attitude. Of all determinants, only six can be pooled for meta-analyses by settings. PA/SB self-efficacy was targeted in the school setting (CTs) and school with family/home setting (RCTs), and there was moderate evidence against an intervention effect. Attitude, subjective norm and parental practice under school with family/home setting (RCTs) also showed moderate evidence against an effect, while the strength of evidence for knowledge cannot be determined. This is surprising as some of these determinants have been widely targeted in PAB/SB interventions [17, 19, 24].

Regarding PAB/SB, results of the meta-analyses showed moderate evidence of absence of post-intervention effect on PAB (CTs) in the school setting and moderate-strong evidence against post-intervention effect on PAB and SB (RCTs) in the school with family/home setting. The certainty of evidence was either low or very low, except for the absence of intervention effect on PAB in the school setting (high certainty). Taken together, the lack of intervention effects on the modifiable determinants might have contributed to unsuccessful PAB/SB change in children.

Considering all studies that showed no corresponding changes between determinants and PAB/SB, we could only conclude that associations between these determinants and PAB/SB were possible, as no change in determinants would not lead to any change in PAB/SB given the assumed association between the targeted determinants and PAB/SB [13]. However, concrete evidence of the associations was not found (i.e. moderate-large effect sizes on both determinants and PAB/SB). Interestingly, we found changes in some determinants but corresponding positive changes in PAB/SB were not evident, namely, perceived barriers to PA at post-intervention [53], self-management and school support at follow-up [53], parental social support [55], external motivation [55], exercise intention [55] and parental modeling in SB [57] (under school and family/home settings) as well as MVPA/SB expectations at follow-up [60] (under community setting). Future research should carefully consider if and how these determinants should be targeted in interventions.

The majority of interventions in the current review were theory-based—as advocated by behaviour change researchers [62]—common theories used being the self-determination theory, goal achievement theory, social cognitive theory and theory of planned behaviour.

However, the intervention effect on determinants and PAB/SB was not significant across settings, despite that most determinants and PAB/SB measurements were evidenced to be psychometrically sound. In fact, our results echoed previous findings on the weak association between some self-determination theory tenets and PAB in the youth population [63]. However, contrary to the results of a recent umbrella reviews of PAB/SB interventions in children, we did not find family support to be associated with behaviour change [64], nor intention or self-efficacy from earlier reviews [26, 27]. Nonetheless, direct comparisons with existing systematic reviews should be cautioned, as the inclusion criteria of different reviews and the analytic strategies are likely to differ. Additionally, the efficacy of theory-based PAB interventions could be compromised by the methodological weakness of the included studies [65], which is potentially applicable to the current review. Whilst every included study inevitably has its limitations, crucially, our results highlighted a bigger picture problem on how the physical inactivity problem is understood, whether it is an individual-level or a population-level issue. Two main factors may have contributed to the failure in changing the determinants and/or (associated) PAB/SB. First and foremost, behaviour change theories that advocate individual-level change solely (including in the context of interpersonal determinants) have long been criticized for their overestimation of people’s self-regulatory ability [66]. Through these theories, individual and interpersonal determinants are derived. A study of Cochrane reviews, from 1993 to 2019, investigating the effect of RCTs on obesity in the youth population shows that consistently, about 57% of interventions target individual and interpersonal determinants only [67]. However, over-reliance on individual agency may have led to failure of policies in tackling the obesity crisis [68] and exacerbated health inequity [69]. This also begs the question–are these determinants as modifiable, or as effective in changing PAB/SB as researchers thought? [14, 15] Perhaps changes in policy and the environment will facilitate change in individual and interpersonal determinants, which will in turn enact the desired behaviour [22]. Such consideration is imperative as different geographical regions have their own PA policies and environmental concerns in place which are likely to affect population-level PAB/SB differently. While there is growing emphasis on targeting policy and environmental determinants through understanding the interactions between actors and determinants within a system (e.g. priority of education policy; a systems-based approach), their changes are challenging to quantify [70]. Nonetheless, realist synthesis can help address the mechanistic associations between the determinants, and enhance our understanding of what works for who, how and in what context [71]. The spirit of realist synthesis can also contain the common problem with heterogeneity in PAB/SB interventions [8, 72]. As such, researchers and public health practitioners should involve stakeholders in developing intervention content specific to a setting that is unique to their needs and political/physical environment [73]. While the systems-based approach might compromise the internal validity of uniform individual-centered interventions, the resulting interventions might see a higher level of buy-in, adherence and ultimately, effectiveness [74].

Another potential explanation for the largely ineffective interventions could be due to an oversight in relatively unconscious motivation that hinders behaviour change [75]. Based on the COM-B model that encapsulates the main ingredients for successful behaviour change [62], all interventions included in this review have targeted individuals’ psychological and/or physical capability (C), have provided social and/or physical opportunities (O), and reflective motivation (M) (referring to the individual/interpersonal psychological determinants), but arguably, automatic motivation to disengage from behaviour change is overlooked. Important to note is that PAB/SB interventions target inactive individuals who are likely to favour being sedentary over being active at a behavioural level [75]. However, PAB promotion seldom considers such inherent resistance to behaviour change [76]. Currently, research into this dual process in behaviour change is largely experimental, so the need for this area of research to be incorporated in applied settings is urgently called for [13]. Not only will such effort benefit intervention design, but also health messaging in the promotion of PAB.

Some limitations of the current review warrant attention. First, this review was based on self-report PAB/SB which is subject to various types of bias, including but not limited to social desirability and recall bias, based on the PAB/SB tools used in the included studies [77]. However, this is not to say that self-reports are inferior to device-based measurements when its usage is fit for purpose [78]. Additionally, the small number of studies included in all meta-analyses makes it challenging to determine the degree of heterogeneity and publication bias, despite that RoBMA was conceptualized to offset the lack of power [43]. Nevertheless, if followed the frequentist approach, publication bias should only be assessed when there are 10 or more studies in a meta-analysis, which all our meta-analyses fall short on [79]. Besides, adopting the Bayesian approach to meta-analysis has benefited our interpretation of the findings, as it can indicate the strength of evidence of the likelihood of the presence or absence of an effect, unlike the all-or-nothing interpretation from the frequentist approach [44]. Another issue with a small study number within a meta-analysis might have contributed to imprecision in the GRADE process even though the total sample is relatively sizable. This has inevitably impacted our assessment of the certainty of evidence. Regarding the risk of bias assessment, as blinding of participants is inherently challenging, if not impossible, due to ethical considerations, the relevant domains related to outcome assessment were deemed ‘high risk’ for all studies. We have thus examined all domains in the risk of bias assessment in the GRADE process instead of relying on the overall risk. Future interventions should consider including an active control group, so that the status of the intervention group can be more easily masked, and any efforts in blinding participants should be made more explicitly clear. Importantly, due to the lack of mediation analyses in the included studies, the association between the determinants and PAB/SB could only be inferred. Whilst the call for mediation analysis to examine the causal pathways was made more than a decade ago [80], many interventions still do not adopt this analytic approach. A potential reason could be the sheer number of determinants (some more modifiable than others) included in some interventions hinder meaningful mediation analyses [81]. For example, in one of the included studies, there are altogether 44 determinants for PAB/SB, and some of these determinants are conceptually similar (e.g., both determinants ‘parents let child watch TV’ and ‘parents remind child about rules’ can fall under one umbrella determinant ‘parental practice on SB’) [57]. Additionally, contradictory evidence exists in the association between determinants and PAB/SB, and researchers ought to monitor their unconscious bias in selecting the determinants to intervene. Without a clearer understanding of the context through which determinants operate and interact with each other, incorporating even evidence-based determinants into an intervention would not guarantee intervention success. Lastly, due to the restrictions of our eligibility criteria, interventions that implemented policy/environment change (as determinants themselves), but without quantifying the magnitude of change, had been excluded from the review. However, these interventions may provide valuable qualitative information regarding the interactions between different levels of determinants within the socio-ecological model. Future research should also review interventions and real-life public health initiatives that targeted policy and environmental change, to examine the extent to which they can effectively modify individual and interpersonal determinants.

Conclusion

The current systematic review set out to examine modifiable determinants in interventions following the RCT and CT design that target children and their association with self-report PAB/SB in different settings. However, the lack in intervention effect on determinants and the corresponding PAB/SB in all settings led us to conclude that the associations between any modifiable determinants and PAB/SB remain uncertain. Specifically, almost all modifiable determinants identified belonged to individual or interpersonal categories according to the socio-ecological model. None of the meta-analyses showed evidence for the presence of intervention effect on the determinants and PAB/SB. These results made us question the modifiability of individual and interpersonal determinants in different settings, and whether they would be more modifiable if policy and/or environment conducive to PAB/SB change were in place. Additionally, for determinants that have seen an intervention effect, but without corresponding changes in PAB/SB, if and how they should be targeted in future interventions should be carefully considered. Crucially, to accelerate our understanding of what determinants might work for who and how, and in what settings, realist synthesis should be conducted in order to inform the design of interventions, and interventions should adopt a system-based approach. With more careful consideration of determinants to target in interventions, conducting mediation analysis between determinants and PAB/SB could provide a clearer picture of their causal pathways. Lastly, design of interventions for children should also consider the automatic motivation that hinders behaviour change.

Supporting information

S1 Checklist. PRISMA checklist.

https://doi.org/10.1371/journal.pone.0309890.s001

(DOC)

S1 File. Effect size and composite score calculation.

https://doi.org/10.1371/journal.pone.0309890.s002

(DOCX)

S2 File. Frequentist analysis.

Frequentist approach to the meta-analyses.

https://doi.org/10.1371/journal.pone.0309890.s003

(DOCX)

S1 Data. Excluded studies with reasons for exclusion.

https://doi.org/10.1371/journal.pone.0309890.s004

(XLSX)

S2 Data. Full data extraction with data extractor and date of data extraction.

https://doi.org/10.1371/journal.pone.0309890.s005

(XLSX)

S1 Appendix. Effect sizes and CIs for all outcomes.

https://doi.org/10.1371/journal.pone.0309890.s006

(DOCX)

References

1. 1. NCD Risk Factor Collaboration (NCD-RisC). Worldwide trends in body-mass index, underweight, overweight, and obesity from 1975 to 2016: a pooled analysis of 2416 population-based measurement studies in 128.9 million children, adolescents, and adults. Lancet. 2017; 390: 2627–42.

* View Article

* Google Scholar

2. 2. Auber S, Barnes JD, Demchenko I, Hawthorne M, Abdeta C, et al. Global matrix 4.0 physical activity report card grades for children and adolescents: Results and analyses from 57 countries. J Phys Act Health. 2022; 19: 700–728. pmid:36280233

* View Article

* PubMed/NCBI

* Google Scholar

3. 3. Morseth B, Jørgensen L, Emaus N, Wilsgaard T. Tracking of leisure time physical activity during 28 yr in adults: The Tromsø Study. Med Sci Sports Exerc. 2011; 43: 1229–34.

* View Article

* Google Scholar

4. 4. Telama R, Yang X, Leskinen E, Kankaanpää A, Hirvensalo M, Tammelin T, et al. Tracking of physical activity from early childhood through youth into adulthood. Med Sci Sports Exerc. 2014; 46: 955–962. pmid:24121247

* View Article

* PubMed/NCBI

* Google Scholar

5. 5. Santos AC, Willumen J, Meheus F, Ilbawi A, Bull FC. The cost of inaction on physical inactivity to public health-care systems: a population-attributable fraction analysis. Lancet Glob Health. 2023; 11: e32–39. pmid:36480931

* View Article

* PubMed/NCBI

* Google Scholar

6. 6. Varela AR, Pratt M, Harris J, Lecy J, Salvo D. Mapping the historical development of physical activity and health research: A structured literature review and citation network analysis. Prev Med. 2018; 111: 466–472. pmid:29709233

* View Article

* PubMed/NCBI

* Google Scholar

7. 7. Tremblay MS, Aubert S, Barnes JD, Saunders TJ, Carson V, Latimer-Cheung AE, et al. Sedentary behavior research network (SBRN)–terminology consensus project process and outcome. Int J Behav Nutr Phys Act. 2017;14:75. pmid:28599680

* View Article

* PubMed/NCBI

* Google Scholar

8. 8. Virgara R, Phillips A, Lewis LK, Baldock K, Wolfenden L, Ferguson T, et al. Interventions in outside-school hours childcare settings for promoting physical activity amongst schoolchildren aged 4 to 12 years. Cochrane Database Syst Rev. 2021; 9: CD013380. pmid:34694005

* View Article

* PubMed/NCBI

* Google Scholar

9. 9. Neil-Sztramko SE, Caldwell H, Dobbins M. School-based physical activity programs for promoting physical activity and fitness in children and adolescents aged 6 to 18. Cochrane Database Syst Rev. 2021; 9: CD007651. pmid:34555181

* View Article

* PubMed/NCBI

* Google Scholar

10. 10. Yuksel HS, Şahin FN, Maksimovic N, Drid P, Bianco A. School-based intervention programs for preventing obesity and promoting physical activity and fitness: A systematic review. Int J Environ Res Public Health. 2020; 17: 347. pmid:31947891

* View Article

* PubMed/NCBI

* Google Scholar

11. 11. Blackburn NE, Wilson JJ, McMullan II, Casserotti P, Giné-Garriga M, Wirth K, et al. The effectiveness and complexity of interventions targeting sedentary behaviour across the lifespan: a systematic review and meta-analysis. Int J Behav Nutr Phys Act. 2020;17:53. pmid:32334631

* View Article

* PubMed/NCBI

* Google Scholar

12. 12. Biddle SJH, Petroliini I, Pearson N. Interventions designed to reduce sedentary behaviours in young people: a review of reviews. Br J Sports Med. 2014; 48: 182–186 pmid:24347578

* View Article

* PubMed/NCBI

* Google Scholar

13. 13. Biddle SJH, Gorely T, Faulkner G, Mutrie N. Psychology of physical activity: a 30-year reflection on correlates, barriers, and theory. Int J Sport Exerc Psychol. 2022; 21: 1–14.

* View Article

* Google Scholar

14. 14. Condello G, Ling FCM, Bianco A, Chastin S, Cardon G, Ciarapica D, et al. Using concept mapping in the development of the EU-PAD framework (EUropean-Physical Activity Determinants across the life course): a DEDIPAC study. BMC Public Health. 2016; 16: 1145. pmid:27825370

* View Article

* PubMed/NCBI

* Google Scholar

15. 15. Chastin SFM, de Craemer M, Lien N, Bernaards C, Buck C, Oppert J-M, et al. The SOS-framework (Systems of Sedentary behaviours): an international transdisciplinary consensus framework for the study of determinants, research priorities and policy on sedentary behaviour across the life course: a DEDIPAC-study. Int J Behav Nutr Phys Act. 2016;13:83. pmid:27421750

* View Article

* PubMed/NCBI

* Google Scholar

16. 16. Sallis JF, Cervero RB, Ascher W, Henderson KA, Kraft MK, Kerr J. An ecological approach to creating active living communities. Annu Rev Public Health. 2006; 27:297–322. pmid:16533119

* View Article

* PubMed/NCBI

* Google Scholar

17. 17. Stierlin AS, de Lepeleere S, Cardon G, Dargent-Molina P, Hoffmann B, Murphy MH, et al. A systematic review of determinants of sedentary behaviour in youth: a DEDIPAC study. Int J Behav Nutr Phys Act. 2015;12:133. pmid:26453175

* View Article

* PubMed/NCBI

* Google Scholar

18. 18. Hu D, Zhou S, Crowley-McHattan ZJ, Liu Z. Factors that influence participation in physical activity in school-aged children and adolescents: A systematic review from the Social Ecological Model perspective. Int J Environ Res Public Health. 2021; 18: 3147. pmid:33803733

* View Article

* PubMed/NCBI

* Google Scholar

19. 19. Cortis C, Puggina A, Pesce C, Aleksovsak K, Buck C, Burns C, et al. Psychological determinations of physical activity across the life course: A “DEterminants of DIet and Physical ACtivity” (DEDIPAC) umbrella systematic literature review. PLoS One. 2017; 12:e0182709.

* View Article

* Google Scholar

20. 20. Ding D, Varela AR, Bauman AE, Ekelund U, Lee I-M, Heath G. et al. Towards better evidence-informed global action: lessons learnt from the Lancet series and recent developments in physical activity and public health. Br J Sports Med. 2020; 54: 462–468. pmid:31562122

* View Article

* PubMed/NCBI

* Google Scholar

21. 21. Pawson R. Evidence-based Policy. A realist perspective. 1^st ed. London: Sage Publications; 2006.

22. 22. Defever E, Jones M. Rapid realist review of school-based physical activity interventions in 7- to 11-year old children. Children. 2021; 8:52. pmid:33467132

* View Article

* PubMed/NCBI

* Google Scholar

23. 23. Brown HE, Atkin AJ, Panter J, Wong G, Chinapaw MJM, van Sluijs EMF. Family-based interventions to increase physical activity in children: a systematic review, meta-analysis and realist synthesis. Obes Rev. 2016; 17:345–360. pmid:26756281

* View Article

* PubMed/NCBI

* Google Scholar

24. 24. Jaeschke L, Steinbrecher A, Luzak A, Puggina A, Aleksovska K, Buck C, et al. Socio-cultural determinants of physical activity across the life course: A ‘Determinants of Diet and Physical Activity’ (DEDIPAC) umbrella systematic literature review. Int J Behav Nutr Phys Act. 2017; 14: 1–15.

* View Article

* Google Scholar

25. 25. Berger KS. The developing person through the lifespan. 12^th ed. Worth Publisher; 2023.

26. 26. Craggs C, Corder K, van Sluijs MF, Griffin SJ. Determinants of change in physical activity in children and adolescents. Am J Prev Med. 2011; 40: 645–658.

* View Article

* Google Scholar

27. 27. Uijtdewilligen L, Nauta j, Singh AS, van Mechelen W, Twisk JWR, van der Horst K, et al. Determinants of physical activity and sedentary behaviour in young people: a review and quality synthesis of prospective studies. Br J Sports Med. 2011; 45: 896–905. pmid:21836173

* View Article

* PubMed/NCBI

* Google Scholar

28. 28. Fiedler J, Eckert T, Burchartz A, Woll A, Wunsch K, et al. Comparison of self-reported and device-based measured physical activity using measures of stability, reliability, and validity in adults and children. Sensors. 2021; 21: 2672. pmid:33920145

* View Article

* PubMed/NCBI

* Google Scholar

29. 29. Colley RC, Butler G, Garriguet D, Prince SA, Roberts KC. Comparison of self-reported and accelerometer-measured physical activity among Canadian youth. Health Rep. 2019; 30: 3–12. pmid:31314124

* View Article

* PubMed/NCBI

* Google Scholar

30. 30. Adamo KB, Prince SA, Tricco AC, Connor-Gorber S, Tremblay M. A comparison of indirect versus direct measures for assessing physical activity in the pediatric population: A systematic review. Int J Pediatr Obes. 2009; 4:2–27. pmid:18720173

* View Article

* PubMed/NCBI

* Google Scholar

31. 31. Moher D, Shamseer L, Clarke M, Ghersi D, Liberati A, et al. Preferred Reporting Items for Systematic Review and Meta-Analysis Protocols (PRISMA-P) 2015 statement. Syst Rev. 2015; 4(1):1 pmid:25554246

* View Article

* PubMed/NCBI

* Google Scholar

32. 32. World Health Organization. Global recommendations on physical activity for health. Geneva: World Health Organisation, 2010.

33. 33. Khudair M, Marcuzzi A, Ng K, Tempest GD, Bartoš F, Peric R, et al. DE-PASS Best Evidence Statement (BESt): modifiable determinants of physical activity and sedentary behaviour in children and adolescents aged 5–19 years–a protocol for systematic review and meta-analysis. BMJ Open. 2022; 12:e059202. pmid:36127107

* View Article

* PubMed/NCBI

* Google Scholar

34. 34. Sterne JAC, Savović J, Page MJ, Elbers RG, Blencowe NS, Boutron I, et al. RoB 2: a revised tool for assessing risk of bias in randomised trials. BMJ. 2019; 366: l4898. pmid:31462531

* View Article

* PubMed/NCBI

* Google Scholar

35. 35. Sterne JAC, Hernán MA, Reeves BC, Savović J, Berkman ND, Viswanathan M, et al. ROBINS-I: a tool for assessing risk of bias in non-randomized studies of interventions. BMJ. 2016; 355; i4919.

* View Article

* Google Scholar

36. 36. McGuinness LA, Higgins JPT. Risk-of-bias VISualization (robvis): An R package and Shiny web app for visualizing risk-of-bias assessments. Res Syn Meth 2020; 1–7. pmid:32336025

* View Article

* PubMed/NCBI

* Google Scholar

37. 37. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, Alonso-Coello P, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ. 2008;336(7650):924–926. pmid:18436948

* View Article

* PubMed/NCBI

* Google Scholar

38. 38. Borenstein M, Hedges LV, Higgins JPT, Rothstein H. Chapter 29: Multiple outcomes or time-points within a study. In: Introduction to meta-analysis. John Wiley & Sons; 2021. Pp.263–276.

39. 39. Maier M, Bartoš F, Wagenmakers E-J. Robust Bayesian meta-analysis: Addressing publication bias with model-averaging. Psychol Methods. 2023; 28: 107–122. pmid:35588075

* View Article

* PubMed/NCBI

* Google Scholar

40. 40. JASP Team. JASP (version 0.16.4), 2021.

41. 41. Bartoš F, Maier M. RoBMA: An R package for robust Bayesian meta-analyses [R package version 2.1. 0]. 2020.

* View Article

* Google Scholar

42. 42. Carpenter B, Gelman A, Hoffman MD, Lee D, Goodrich B, Betancourt M, et al. Stan: A probabilistic programming language. J Stat Softw. 2017; 76: 1–32. pmid:36568334

* View Article

* PubMed/NCBI

* Google Scholar

43. 43. Bartoš F, Maier M, Wagenmakers E-J, Doucouligagos H, Stanley TD. Robust Bayesian meta-analysis: Model-averaging across complementary publication bias adjustment methods. Res Synth Methods. 2023; 14: 99–116. pmid:35869696

* View Article

* PubMed/NCBI

* Google Scholar

44. 44. Lee MD, Wagenmakers E-J. Bayesian cognitive modeling: A practical course. Cambridge: Cambridge University Press; 2013.

45. 45. Cohen J. Statistical power analysis for the behavioral sciences. New York: Routledge Academic; 1988.

46. 46. Vevea JL, Hedges LV. A general linear model for estimating effect size in the presence of publication bias. Psychometrika. 1995; 60:419–435.

* View Article

* Google Scholar

47. 47. Boyle-Holmes T, Grost L, Russell L, Laris BA, Robin L, Haller E, et al. Promoting elementary physical education: Results of a school-based evaluation study. Health Educ Beh. 2010; 37:377–389. pmid:19749086

* View Article

* PubMed/NCBI

* Google Scholar

48. 48. Gråstén A, Yli-Piipari S. The patterns of moderate to vigorous physical activity and physical education enjoyment through a 2-year school-based program. J Sch Health. 2019; 89:88–98. pmid:30604450

* View Article

* PubMed/NCBI

* Google Scholar

49. 49. Gabriel KKP, DeBate RDG, High RR, Racine EF. Girls on the Run: A quasi-experimental evaluation of a developmentally focused youth sport program. J Phys Act Health. 2011; 8: S285–S294.

* View Article

* Google Scholar

50. 50. Lonsdale C, Lester A, Owen KB, White RL, Moyes I, Peralta L, et al. An internet-supported school physical activity intervention in low socioeconomic status communities: results from the Activity and Motivation in Physical Education (AMPED) cluster randomised controlled trial. Br J Sports Med. 2019; 53:341–347. pmid:28993404

* View Article

* PubMed/NCBI

* Google Scholar

51. 51. Wang JJ, Baranowski T, Lau PWC, Buday R, Gao Y. Story immersion may be effective in promoting diet and physical activity in Chinese children. J Nutr Educ Behav. 2017; 49:321–329. pmid:28391798

* View Article

* PubMed/NCBI

* Google Scholar

52. 52. Maddison R, Marsh S, Foley L, Epstein LH, Olds T, Dewes O, et al. Screen-time Weight-loss Intervention Targeting Children at Home (SWiTCH): a randomized controlled trial. Int J Beh Nutr Phys Act. 2014; 11:111. pmid:25204320

* View Article

* PubMed/NCBI

* Google Scholar

53. 53. Pearce K, Dollman J. Healthy for Life pilot study: A multicomponent school an dhome based physical activity intervention for disadvantaged children. Int J Environ Res Public Health. 2019; 16:2935. pmid:31443294

* View Article

* PubMed/NCBI

* Google Scholar

54. 54. Bergh IH, van Stralen MM, Bjelland M, Grydeland M, Lien N, Klepp K-I, et al. Post-intervention effects on screen behaviours and mediating effect of parental regulation: the HEalth In Adolescents study–a multi-component school based randomized controlled trial. BMC Public Health. 2014; 14:200. pmid:24568125

* View Article

* PubMed/NCBI

* Google Scholar

55. 55. Quaresma AM, Palmeira AL, Martins SS, Minderico CS, Sardinha LB. Effect of a school-based intervention on physical activity and quality of life through serial mediation of social support and exercise motivation: the PESSOA program. Health Educ Res. 2014; 29:906–917. pmid:25274722

* View Article

* PubMed/NCBI

* Google Scholar

56. 56. Zhang Y, Yin Y, Liu J, Yang M, Liu Z, Ma X. Impact of combined theory-based intervention on psychological effects and physical activity among Chinese adolescents. Int J Environ Res Public Health. 2020; 17:3026. pmid:32349260

* View Article

* PubMed/NCBI

* Google Scholar

57. 57. Vik FN, Lien N, Berntsen S, De Bourdeaudhuij I, Grillenberger M, Manios Y, et al. Evaluation of the UP4FUN intervention: A cluster randomized trial to reduce and break up sitting time in European 10–12 year-old children. PLoS One. 2015; 10:31022612. pmid:25826704

* View Article

* PubMed/NCBI

* Google Scholar

58. 58. Moitra P, Madan J, Verma P. Impact of a behaviourally focused nutrition education intervention on attitudes and practices related to eating habits and activity levels in Indian adolescents. Public Health Nutr. 2019; 24: 2715–26.

* View Article

* Google Scholar

59. 59. Salmon J, Jorna M, Hume C, Arundell L, Chahine N. A translational research intervention to reduce screen behaviours and promote physical activity among children: Switch-2-Activity. Health Promot Int. 2011; 26:311–321. pmid:21177770

* View Article

* PubMed/NCBI

* Google Scholar

60. 60. Branscum P, Sharma M, Wang LL, Wang LL, Wilson B, Rojas-Guyler L. A true challenge for any superhero: An evaluation of a comic book obesity prevention program. Fam Comm Health. 2013; 36:63–76. pmid:23168347

* View Article

* PubMed/NCBI

* Google Scholar

61. 61. Christiansen LB, Toftager M, Ersbøll A, Troelssen J. Effects of a Danish multicomponent physical activity intervention on active school transport. J Transport Health. 2014; 1:174–181.

* View Article

* Google Scholar

62. 62. Michie S, van Stralen MM, West R. The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implement Sci. 2011; 6:42. pmid:21513547

* View Article

* PubMed/NCBI

* Google Scholar

63. 63. Owen KB, Smith J, Lubans DR, Ng JYY, Lonsdale C. Self-determined motivation and physical activity in children and adolescents: A systematic review and meta-analysis. Prev Med. 2014; 67:270–279. pmid:25073077

* View Article

* PubMed/NCBI

* Google Scholar

64. 64. Messing S, Rütten A., Abu-Omar K, Ungerer-Röhrich U, Goodwin L, Burlacu I, et al. How can physical activity be promoted among children and adolescents? A systematic review of reviews across settings. Frontiers in Public Health. 2019; 7:55. pmid:30941342

* View Article

* PubMed/NCBI

* Google Scholar

65. 65. Bernard P, Carayol M, Gourlan M, Boiché J, Romain AJ, Bortolon C, et al. Moderators of theory-based interventions to promote physical activity in 77 randomized controlled trials. Health Edu Beh. 2017; 44: 227–235. pmid:27226432

* View Article

* PubMed/NCBI

* Google Scholar

66. 66. Davis R, Campbell R, Hildon Z, Hobbs L, Michie S. Theories of behaviour and behaviour change across the social and behavioural sciences: a scoping review. Health Psych Rev. 2015; 9:323–344. pmid:25104107

* View Article

* PubMed/NCBI

* Google Scholar

67. 67. Nobles J, Summerbell C, Brown T, Jago R, Moore T. A secondary analysis of the childhood obesity prevention Cochrane Review through a wider determinants of health lens: implications for research funders, researchers, policymakers and practitioners. Int J Beh Nutr Phys Act. 2021; 18:22. pmid:33563281

* View Article

* PubMed/NCBI

* Google Scholar

68. 68. Theis DRZ, White M. Is obesity policy in England fit for purpose? Analysis of government strategies and policies, 1992–2020. The Milbank Quarterly. 2021; 99: 126–170. pmid:33464689

* View Article

* PubMed/NCBI

* Google Scholar

69. 69. Adams J, Mytton O, White M, Monsivais P. Why are some population interventions for diet and obesity more equitable and effective than others? The role of individual agency. PLoS Med. 2016; 13:e1001990. pmid:27046234

* View Article

* PubMed/NCBI

* Google Scholar

70. 70. Rutter H, Savona N, Glonti K, Bibby J, Cummins S, Finegood DT. The need for a complex systems model of evidence for public health. Lancet. 2017; 390:2602–2604. pmid:28622953

* View Article

* PubMed/NCBI

* Google Scholar

71. 71. Rycroft-Malone J, McCormack B, Hutchinson AM, DeCorby K, Bucknall TK, Kent B, et al. Realist synthesis: illustrating the method for implementation research. Implement Sci. 2012; 7:33. pmid:22515663

* View Article

* PubMed/NCBI

* Google Scholar

72. 72. Bleich SN, Segal J, Wu Y, Wilson R, Wang Y. Systematic review of community-based childhood obesity prevention studies. Pediatrics. 2013; 132:e201–210. pmid:23753099

* View Article

* PubMed/NCBI

* Google Scholar

73. 73. Brandes M, Brandes B, Sell L, Sacheck JM, Chinapaw M, Lubans DR, et al. How to select interventions for promoting physical activity in schools? Combining preferences of stakeholders and scientists. Int J Beh Nutr Phys Act. 2023; 20:48.

* View Article

* Google Scholar

74. 74. Jago R, Salway R, House D, Beets M, Lubans DR, Woods C, et al. Rethinking children’s physical activity interventions at school: A new context-specific approach. Front Public Health. 2023; 11:1149883. pmid:37124783

* View Article

* PubMed/NCBI

* Google Scholar

75. 75. Wang Y-B, Sun X-T, Mao Z-X. Is approaching sedentary behaviour or physical activity a reward? An EEG study. Beh Brain Res. 2023; 445.

* View Article

* Google Scholar

76. 76. Brand R, Cheval B. Theories to explain exercise motivation and physical inactivity: Ways of expanding our current theoretical perspective. Front Psychol. 2019; 10:1147. pmid:31164856

* View Article

* PubMed/NCBI

* Google Scholar

77. 77. Hidding LM, Chinapaw MJ, van Poppel MN, Mokkink LB, Altenburg TM. An updated systematic review of childhood physical activity questionnaires. Sports Med. 2018; 48:2797–842. pmid:30298479

* View Article

* PubMed/NCBI

* Google Scholar

78. 78. Sattler MC, Ainsworth BE, Andersen LB, Foster C, Hagströmer M, Jaunig J, et al. Physical activity self-reports: past or future? Br J Sports Med 2021; 55: 889–890. pmid:33536193

* View Article

* PubMed/NCBI

* Google Scholar

79. 79. Sterne JAC, Sutton AJ, Ioannidis JPA, Terrin N, Jones DR, Lau J, et al. Recommendations for examining and interpreting funnel plot asymmetry in meta-analyses of randomised controlled trials. BMJ. 2011; 343:d4002. pmid:21784880

* View Article

* PubMed/NCBI

* Google Scholar

80. 80. van Sluijs MF, Kriemler S, McMinn AM. The effect of community and family interventions on young people’s physical activity levels: a review of reviews and updated systematic review. Br J Sports Med. 2011; 45:914–922. pmid:21836175

* View Article

* PubMed/NCBI

* Google Scholar

81. 81. Agler R, De Boeck P. On the Interpretation and Use of Mediation: Multiple Perspectives on Mediation Analysis. Front Psychol. 2017; 8:1984. pmid:29187828

* View Article

* PubMed/NCBI

* Google Scholar

Citation: Ling FCM, Khudair M, Ng K, Tempest GD, Peric R, Bartoš F, et al. (2024) DE-PASS Best Evidence Statement (BESt): Determinants of self-report physical activity and sedentary behaviours in children in settings: A systematic review and meta-analyses. PLoS ONE 19(11): e0309890. https://doi.org/10.1371/journal.pone.0309890

About the Authors:

Fiona C. M. Ling

Contributed equally to this work with: Fiona C. M. Ling, Mohammed Khudair, Kwok Ng, Gavin D. Tempest, Ratko Peric, František Bartoš, Maximilian Maier, Laura Capranica, Ciaran MacDonncha, Anna Marcuzzi

Roles: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Writing – original draft, Writing – review & editing

E-mail: [email protected]

Affiliation: Department of Sport, Exercise and Rehabilitation, Northumbria University, Newcastle Upon Tyne, United Kingdom

ORICD: https://orcid.org/0000-0003-2357-3779

Mohammed Khudair

Roles: Data curation, Formal analysis, Investigation, Methodology, Writing – review & editing

Affiliation: Department of Sport, Exercise and Rehabilitation, Northumbria University, Newcastle Upon Tyne, United Kingdom

Kwok Ng

Roles: Formal analysis, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing

Affiliations: Department of Education, University of Turku, Rauma, Finland, Department of Physical Education and Sport Sciences, University of Limerick, Limerick, Ireland, Institute of Innovation and Sports Science, Lithuanian Sport University, Kaunas, Lithuania

Gavin D. Tempest

Roles: Investigation, Methodology, Writing – original draft, Writing – review & editing

Affiliation: Department of Sport, Exercise and Rehabilitation, Northumbria University, Newcastle Upon Tyne, United Kingdom

Ratko Peric

Roles: Investigation, Methodology, Writing – review & editing

Affiliation: Exercise Physiology Laboratory, OrthoSport Banja Luka, Banja Luka, Bosnia

ORICD: https://orcid.org/0000-0003-3345-6293

František Bartoš

Roles: Formal analysis, Methodology, Writing – original draft, Writing – review & editing

Affiliation: Department of Psychological Methods, University of Amsterdam, Amsterdam, Netherlands

ORICD: https://orcid.org/0000-0002-0018-5573

Maximilian Maier

Roles: Formal analysis, Methodology, Writing – original draft, Writing – review & editing

Affiliation: Department of Experimental Psychology, University College London, London, United Kingdom

Mirko Brandes

Roles: Investigation, Writing – review & editing