1 Introduction

Physical inactivity among children and adolescents has become a persistent global public health concern rather than a transient lifestyle issue. Worldwide surveillance indicates that approximately 81% of adolescents aged 11–17 years fail to achieve the World Health Organization (WHO) recommendation of at least 60 minutes of moderate-to-vigorous physical activity (MVPA) daily [1,2]. Despite sustained policy attention and large-scale school- and community-based initiatives, population-level physical activity (PA) levels in children and adolescents have shown little improvement over the past decade [2,3]. This pattern is concerning given the well-established associations between insufficient PA and adverse cardiometabolic, mental health, and cognitive outcomes in childhood and adolescence [3–5]. Moreover, PA behaviors track strongly from adolescence into adulthood, contributing to the long-term burden of non-communicable diseases across the life course [5,6].

In response, numerous PA promotion interventions targeting children and adolescents have been developed. However, recent systematic reviews and meta-analyses consistently report small and heterogeneous effects, with limited evidence of sustained behavior change [3,4,7]. One frequently cited explanation is the limited use of behavioral theory in intervention design and reporting [8]. Many interventions focus on educational or environmental components without clearly articulating the behavioral determinants they aim to modify or the mechanisms through which change is expected to occur [7,9]. This “black box” approach constrains cumulative learning, impedes replication, and weakens translation into practice.

Contemporary guidance for complex interventions, including that from the UK Medical Research Council, emphasizes the importance of theory-informed development to support more transparent intervention development, evaluation, and interpretation [10,11]. Within this context, the Behavior Change Wheel (BCW) has emerged as a widely used integrative framework for intervention design [12]. The BCW is underpinned by the Capability, Opportunity, and Motivation-Behavior (COM-B) model as its core, which frames behavior as a function of capability, opportunity, and motivation. More specifically, the BCW is organized into three interrelated layers: at the center is the COM-B model, which provides a framework for behavioral diagnosis; surrounding this are nine intervention functions that guide the selection of broad strategies for changing behavior; and the outer layer comprises policy categories that can support intervention delivery and wider implementation [12]. This structure supports a systematic translation from behavioral diagnosis to the selection of intervention functions and the specification of intervention content using behavior change techniques (BCTs) [12,13]. By linking behavioral analysis to intervention planning in this way, the BCW helps make the theoretical basis of an intervention more explicit and facilitates more transparent reporting of how intervention content is derived from identified determinants of behavior [12]. Over the past decade, the BCW has been applied across a range of health behaviors and settings, including smoking cessation, diet, diabetes management, antimicrobial stewardship, and PA promotion [14–17]. This may be especially relevant for PA promotion among children and adolescents, as PA behavior in these populations is shaped by interacting individual, interpersonal, institutional, and environmental factors.

Nevertheless, important gaps remain in how the BCW has been applied to PA promotion among children and adolescents. This population’s PA is developmentally contingent and strongly conditioned by structural opportunity, including school routines, parental support, and neighborhood resources. These features complicate behavioral diagnosis and raise questions about whether COM-B constructs are being operationalized in ways that reflect the realities of this population’s settings and decision-making [9,18]. At the same time, there is a recurring concern that the BCW is sometimes invoked more as a marker of rigor than as a framework that is used systematically across the full design pathway, a practice often described as theoretical tokenism [19]. Where this occurs, the mapping process is frequently left opaque. In particular, the linkages from COM-B components to BCW intervention functions and then to specific BCTs are not consistently reported, making it difficult to identify the active ingredients and to accumulate learning across studies [17,19]. Moreover, the literature continues to privilege effectiveness outcomes, while reporting of implementation processes such as feasibility, acceptability, and fidelity remains limited, which constrains inference about how BCW-informed interventions are implemented, reported, and potentially sustained in real-world youth contexts [20,21].

To date, no scoping review has systematically mapped how the BCW has been applied to promote PA among children and adolescents, nor examined the transparency of theoretical mapping and reporting in this field. A scoping review is particularly appropriate for this purpose because the key question is not whether BCW-informed interventions are more effective than other theory-based interventions, but what is known about how the BCW and COM-B model have been used, operationalized, and reported in this area. Addressing this gap is critical for strengthening theory-informed PA interventions and supporting clearer reporting, replication, and future intervention development. Therefore, this scoping review aims to: (1) map the extent and characteristics of BCW applications in PA promotion among this population; (2) examine how behavioral determinants, intervention functions, and BCTs are selected and reported; and (3) summarize reported implementation outcomes and identify gaps in theoretical mapping and reporting practices. By clarifying how behavioral theory is currently operationalized, this review seeks to inform future theory-informed development and reporting of PA interventions for children and adolescents.

2. Methods

3 Results

3.1 Selection of sources of evidence

The study selection process is illustrated in Fig 1. Database searches identified 743 records. After removal of duplicates (n = 152) and records with insufficient bibliographic information (n = 78), 513 records remained for title and abstract screening. Of these, 456 records were excluded as clearly not meeting the inclusion criteria. Fifty-seven full-text articles were assessed for eligibility. Following full-text review, 42 articles were excluded, primarily because they were review papers (n = 9), did not report relevant empirical data (n = 19), or did not involve the target population of children or adolescents (n = 14). Ultimately, 15 studies met all eligibility criteria and were included in the scoping review [18,25–38].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Flowchart of the literature search process.

https://doi.org/10.1371/journal.pone.0354697.g001

3.2 Characteristics of included studies

The key characteristics of the included studies are summarized in Table 1. The studies were published between 2015 and 2025 and were conducted predominantly in Ireland [18,25,27,32–34], followed by the United Kingdom [28,29,36], Australia [30,35], China [37,38], the United States [26], and Belgium [31].

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of included studies.

https://doi.org/10.1371/journal.pone.0354697.t001

Most interventions were implemented in school-based settings [18,25,27,29,31–33,35–38], often supplemented by family [26,34,37], community [28,33–35,37], or digital components [25,26,28]. Several studies targeted specific clinical or vulnerable populations, including childhood cancer survivors [26], children with cerebral palsy [35], and adolescents with intellectual disabilities [31,33], while others focused on general school-aged populations [18,25,27–30,32,34,36–38].

In terms of study design, a substantial proportion of the literature focused on intervention development, co-design, feasibility, or protocol reporting, rather than definitive effectiveness trials [18,25,30–32,34,37,38]. Accordingly, PA outcomes were heterogeneous. Some studies employed objective or combined measurement approaches (e.g., accelerometry, pedometers) [26,27,29,32], whereas others relied on self-report measures or did not report behavioral outcomes due to their developmental focus [18,25,28,30,31,33–38].

The use of digital technologies and wearable devices (e.g., Fitbit-based interventions) was common [18,25–31,35,37,38], reflecting a growing emphasis on technology-supported behavior change in PA interventions of children and adolescents.

3.3 Applications of the behavior change wheel and COM-B frameworks

The roles and modes of application of the BCW and COM-B frameworks are detailed in Table 2. Table 2 summarizes how the BCW and COM-B frameworks were applied across studies, including their application role, research stage, the Theory, Model, and Framework Comparison and Selection Tool (T-CaST) level, mapping transparency from COM-B to BCW and BCTs, and co-frameworks used. T-CaST is an implementation science tool for comparing and selecting theories, models, and frameworks. Across the included studies, these frameworks were used predominantly to inform intervention design and development, particularly within co-design and feasibility-focused research. A smaller number of studies applied COM-B for determinant analysis or measurement development.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Use of the behavior change wheel and COM-B.

https://doi.org/10.1371/journal.pone.0354697.t002

3.3.1 Role of frameworks and mapping transparency.

Most studies demonstrated high levels of theoretical integration, explicitly linking COM-B components to BCW intervention functions and, in turn, to specific BCTs. However, several studies reported partial or implicit use of the frameworks, with limited transparency regarding the mapping process.

3.3.2 Use of Co-frameworks.

The BCW and COM-B were frequently combined with other frameworks to enhance analytical depth. The TDF was the most commonly used supplementary framework, followed by the MRC framework for complex interventions, the CALO-RE taxonomy, and evaluative frameworks such as APEASE. This multimodal theoretical approach was particularly evident in studies focused on intervention development and optimization.

3.4 Theoretical mapping of behavioral determinants and intervention functions

The distribution of targeted COM-B components and BCW intervention functions is presented in Fig 2, with detailed mappings provided in S3 File.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Distribution of targeted COM-B components and BCW intervention functions.

https://doi.org/10.1371/journal.pone.0354697.g002

3.4.1 Targeting COM-B components.

All six COM-B components were addressed across the included studies. Psychological capability and reflective motivation were the most frequently targeted determinants, appearing in 87% of studies, indicating a strong emphasis on knowledge, beliefs, and intentional processes related to PA. Physical capability and automatic motivation were targeted in 80% of studies, reflecting the importance of skill development, habit formation, and affective processes in PA promotion of this population. Social and physical opportunity were also frequently addressed (80%), underscoring the role of environmental and social contexts, particularly within school and family settings.

3.4.2 Distribution of BCW intervention functions.

As shown in Fig 2B, the selection of BCW intervention functions was highly concentrated. Education, persuasion, training, modelling, environmental restructuring, and enablement were the most frequently employed functions. These functions aligned closely with the dominant COM-B determinants identified across studies.

In contrast, restriction and coercion were rarely used (7%), and in some cases explicitly excluded, suggesting a consistent preference for autonomy-supportive and developmentally appropriate strategies in interventions targeting children and adolescents.

3.5 Intervention content: Behavior change techniques

The BCTs used across interventions are summarized in Table 3 and illustrated in Fig 3. Table 3 summarizes the BCT-related information reported across studies, including whether BCT mapping was presented, the taxonomy used, the number of BCTs identified, the main BCT clusters, which refer to the 16 higher-order groupings of BCT Taxonomy version 1 (BCTTv1), and the specific BCTs reported.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Applications of behavior change techniques.

https://doi.org/10.1371/journal.pone.0354697.t003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Mapping the applications of behavior change techniques.

https://doi.org/10.1371/journal.pone.0354697.g003

3.5.1 Volume and taxonomy of behavior change techniques.

Among studies that reported explicit BCT mapping, the number of techniques ranged from 11 to 32 per intervention. The majority employed BCT Taxonomy v1 (BCTTv1), facilitating standardized reporting and comparison across studies. A smaller number of studies used earlier taxonomies or ontology-based classifications.

3.5.2 Most frequent behavior change techniques.

As shown in Fig 3B, Goal setting (behavior) [BCT 1.1] was the most consistently applied technique, appearing in 100% of mapped studies (n = 10). A cluster of BCTs was used in 90% of studies, including Action planning (1.4), Social support (unspecified) (3.1), Information about health consequences (5.1), and Instruction on how to perform the behavior (4.1). Additional techniques appearing in ≥80% of studies included Self-monitoring of behavior (2.3), Feedback on behavior (2.2), Demonstration of behavior (6.1), Prompts/cues (7.1), and Verbal persuasion about capability (15.1). Overall, intervention content was characterized by a strong emphasis on self-regulation strategies, complemented by instructional and social support components.

4. Discussion

This scoping review mapped the extent and nature of applications of the BCW framework in PA promotion research for children and adolescents. The included evidence base remains relatively small and skewed toward intervention development, co-design, feasibility and diagnostic work, rather than fully powered effectiveness trials. This pattern is not necessarily a weakness, as recent guidance on complex interventions emphasizes that iterative development, explicit theorization, and early contextual attention are essential prerequisites for large-scale evaluation and scale-up [11].

A distinct “theory-to-technique” profile emerged. In studies offering sufficient detail, intervention content predominantly focused on capability building and self-regulation through strategies such as goal setting, action planning, instruction, feedback, and practical social support. Such clustering is broadly consistent with the wider PA and digital behavior change literature, where planning/monitoring and instructional components are frequently deployed as the core active ingredients [31,39,40]. Although some interventions incorporated interpersonal elements such as practical social support, their overall emphasis remained primarily on capability building and student-level self-regulation. School-based PA scholarship increasingly suggests that such approaches alone rarely shift overall PA levels sustainably because PA opportunities among children and adolescents are structurally bounded by timetables, staff capacity, facilities, safety norms, and competing educational priorities [41–43]. In this sense, the mapped BCT profile may reflect both a strength (a coherent, replicable toolkit) and a potential constraint (a tendency to default to student-level self-regulation even when opportunity constraints are dominant). The practical implication is that the BCW should be utilized not only to justify the selection of specific techniques, but also to ensure that intervention functions effectively address the primary constraints on this population’s PA, which frequently involve physical and social opportunities within school and community settings.

A particularly significant yet overlooked finding in this review is the minimal reliance on BCW functions centered on control or punishment. Specifically, coercion [27] and restriction [30] both appeared in only 7% of studies, meaning that these intervention functions were included in only a few intervention designs and were not used in most studies. This is notable because it may reflect a preference for non-coercive and autonomy-supportive approaches in PA promotion for children and adolescents [44]. The BCW explicitly includes coercion and restriction as potential intervention functions, but their use carries practical and normative risks in this population’s settings (e.g., undermining autonomy, reducing enjoyment, and provoking resistance) [12]. The broader motivational literature supports this interpretation: autonomy-supportive climates are consistently associated with more adaptive motivational profiles and higher PA engagement in adolescents, whereas controlling approaches tend to predict poorer motivation and disengagement [45–47]. For practitioners, this finding reinforces a concrete design principle: youth PA interventions should prioritize positive, competence-building, autonomy-supportive strategies (e.g., choice, mastery-oriented goals, supportive feedback, peer modelling, enjoyable routines), rather than punitive contingencies or restrictive controls. Evidence from self-determination theory-based research indicates that PA intervention effects are often driven by enhanced autonomous motivation and supportive social contexts [45,46]. Such mechanisms are functionally consistent with BCW strategies like enablement and environmental restructuring, yet they remain fundamentally incompatible with coercion-based methods.

Concurrently, the mapped evidence reveals that contextual opportunity levers, particularly those necessitating organizational or systemic change, remain less consistently operationalized and evaluated. Advocacy for rethinking school-based PA interventions suggests that research should integrate with school systems, leverage existing provisions, and establish context-specific causal pathways [41–43]. This involves prioritizing changes in timetables, staff practices, and school culture instead of merely layering new program elements onto already overburdened systems. Policy-oriented frameworks reinforce this systemic logic. For instance, the ISPAH “Eight Investments” advocate for whole-of-school programs, active transport, urban design, and supportive public education as essential interventions to reshape opportunity structures and mitigate inequities [48]. Within this context, the mapped evidence reveals a trend where the BCW serves to define intervention components rather than to reshape delivery environments. Specifically, there is a missed opportunity to leverage the framework for creating the environmental conditions necessary to make PA more accessible and socially normative for young people.

While implementation and sustainability represent the most critical evidence gaps from a public health standpoint [21], these concerns should not overshadow the fundamental challenge of promoting this population’s PA. From a practical perspective, the absence of data on reach, adoption, fidelity, and adaptation makes it nearly impossible for schools, communities, and clinical services to determine whether BCW-informed approaches can be effectively scaled and maintained within routine environments [49]. This gap mirrors broader school health implementation research, where interventions may be efficacious but often fail at the “delivery” stage [50]. A pragmatic approach involves embedding a focused, decision-relevant implementation dataset into PA research. This includes tracking participation and reach, delivery fidelity, and indicators of sustainment potential, while maintaining a primary focus on PA as the effectiveness outcome. Actionable strategies derived from school-based evidence syntheses, such as professional training and the integration of interventions into school workflows, align closely with BCW-derived diagnoses of delivery-system barriers [50].

Digital and wearable components appeared frequently across the included studies, reflecting contemporary trends in PA promotion of children and adolescents. Although digital platforms can amplify key BCTs, they frequently introduce “feature creep” through an over-reliance on multiple functions with unclear mechanisms. When coupled with issues of unequal access and engagement decay, these factors underscore a critical concern frequently emphasized in digital behavior change syntheses [40,51]. The BCW mapping process offers a more rigorous approach in this context. Interventions become more interpretable and easier to optimize when digital features are explicitly justified against specific COM-B determinants, such as linking prompts to environmental opportunity and social features to social opportunity. In parallel, stronger reporting standards are needed so that digital and multi-component PA interventions can be replicated and meaningfully compared. The Template for Intervention Description and Replication (TIDieR) checklist, along with its specialized extensions for complex interventions, offers a practical framework to enhance reporting granularity without compromising the brevity required for academic publication [52,53].

Several priorities follow for the next phase of BCW work in PA promotion of children and adolescents. First, research must transcend simple “component mapping” to establish testable, developmentally plausible causal logic. For instance, if modelling and social support are selected to address social opportunity, evaluations should incorporate intermediate mediators such as perceived peer norms, social inclusion, and enjoyment—constructs empirically linked to adolescent PA participation [45]. Second, outcome measurement should align with the behavioral target and setting constraints. The school-based interventions may change activity patterns within school hours but not necessarily overall MVPA, so selecting outcomes that reflect intended mechanisms (e.g., participation, in-school MVPA, active transport) can reduce false negatives and improve interpretability [41]. Finally, intervention functions and techniques must be selected with explicit consideration of equity. Since this population’s PA is heavily patterned by socioeconomic status, gender, disability, and neighborhood opportunity, integrating meaningful youth co-design can minimize the discrepancy between intervention assumptions and the lived experiences of participants [19,54].

This review provides a structured map of BCW applications specifically within child and adolescent PA promotion, integrating evidence across settings and study designs. By tracing patterns across COM-B determinants, intervention functions and BCT content, the review highlights both a coherent “default toolkit” and where the field may be underleveraging opportunity-focused and system-compatible strategies.

A key limitation of this review is that it focused exclusively on studies explicitly using the BCW and/or COM-B framework. Grey literature, including government reports, conference abstracts, and theses/dissertations, was not included; therefore, some unpublished, practice-based, or policy-oriented applications of the BCW/COM-B model may have been missed. While this enabled a theory-specific synthesis, it may have excluded relevant studies using other theoretical or implementation frameworks that address similar constructs. In addition, the number of eligible studies was relatively limited, particularly across some populations and application contexts, which constrained the breadth of cross-study comparisons. Incomplete reporting of intervention content and mapping pathways in primary studies limited the granularity of cross-study comparisons, reflecting a broader reporting challenge in behavior change research.

5. Conclusion

While the application of the BCW framework in PA promotion of children and adolescents is expanding, research remains predominantly focused on developmental and feasibility stages. In cases where interventions were clearly specified, the prevailing approach integrated goal setting and action planning with instruction, demonstration, and social support. This clusters effectively around capability building and self-regulation. Significantly, the marked absence of coercive or restrictive functions reinforces the principle that PA promotion of this population must prioritize autonomy-supportive strategies over punitive or controlling methods. Moving forward, the BCW should be leveraged to address structural opportunity constraints within schools and communities more directly. Future research must specify mechanisms that align with developmental realities and incorporate a core set of implementation indicators to facilitate real-world adoption and sustainment, ensuring these efforts complement rather than displace the fundamental goal of improving this population’s PA.

Supporting information

Acknowledgments

The authors would like to thank the researchers of the included studies for their foundational work, which provided the essential evidence base for this review.