Formative research to develop a school-based, community-linked physical activity role model programme for girls: CHoosing Active Role Models to INspire Girls (CHARMING)

Globally, an estimated 3.2 million deaths are attributed to physical inactivity, placing physical inactivity as the 4th leading cause of death worldwide [1]. Alongside adverse physiological outcomes such as increased risk of being overweight or obese, and developing type 2 diabetes [2], physical inactivity is also associated with poor mental wellbeing [3]. With three in four mental illnesses starting before age 15 years [4] and over 80% of adolescents currently insufficiently active [5], there is an urgent need to implement effective strategies to increase physical activity (PA) before PA-related health disparities become entrenched. Adolescent girls in particular have low PA levels, displaying a notable decline during the transition period from childhood to adolescence [6]. As a result, girls tend to be far less active than boys of the same age, a trend observed worldwide [7].

Previous work has shown confusion surrounding children’s perceptions of PA requirements [8]. Among a sample of English preadolescents aged 8–12 years, children were able to identify foods comprising a healthy diet, yet they had limited knowledge of PA recommendations [8]. Preadolescence therefore presents a window of opportunity to intervene on both exposures (e.g., unhealthy foods) and behaviours (e.g., sedentary behaviours) that can have lasting health effects [9] and educate children on healthy lifestyles [10]. A large number of PA interventions include health education as an intervention component; however, one review highlighted the uncertainty about how children’s knowledge of (or beliefs and attitudes towards) PA affects their PA participation [11]. Supporting this assertion, a review on the effectiveness of classroom-based education [12] found increases in general health knowledge, exercise-related knowledge, and self-efficacy about exercise lead to limited success in increasing time children spent in PA outside the school setting. To our knowledge, no intervention to date has attempted to deliver health education messages using PA as the mode of delivery, presenting a potential opportunity to bridge the gap between health education and health behaviours.

Schools have become a popular setting for intervention delivery [13]. A recent systematic review [14] examining the impact of school-based PA interventions targeting older girls (e.g. adolescents) highlighted that multi-component interventions underpinned by theory were more likely to be effective. Such findings echo recommendations outlined within the Medical Research Council [MRC] framework when developing complex interventions [15], emphasising the development stage as key. The importance of co-producing health-behaviour interventions with the target audience [16, 17] has been recognised as one strategy to enhance intervention optimisation [18‐21]. To date however, interventions have typically adopted a top-down approach, whereby a researcher engages with the target audience for feedback on developed elements [16] as opposed to undertaking formative work with the target audience during the pre-trial stage [17]. Co-production, a means of involving the target audience in both the design and implementation stages of an intervention, has shown a greater degree of intervention engagement [18], providing the opportunity to co-create a contextually appropriate intervention design. Wight and colleagues [19] emphasise the value of co-production in maximising the likelihood of intervention effectiveness by improving intervention fit with considerations for the target audience perceived needs and acceptability, practicality, intervention theory, and implementation and uptake by stakeholders. Alongside the involvement of a target population, it is also important to engage their gatekeepers (e.g., schools, teachers and parents) in addressing contextual complexities [20]. The necessity of stakeholder involvement throughout intervention design and implementation has been recognised as a crucial element for the facilitation of long-term improvements in PA, sedentary and dietary behaviours [21].

With a specific focus on increasing PA amongst girls, research has shown the importance of providing a wide range of appealing activities [22], providing girl-only sessions [23, 24] and underpinning an intervention framework with theory [15]. Recognising the importance of motivation for the long-term adoption of a behaviour [25], many interventions focus on attempts to promote action by converting motivation into action [26]. Such motivation can come from observing positive results in others and it has been suggested that role models can provide inspiration for young people to become involved in, or maintain involvement in, sport and PA [27]. The World Health Organisation (WHO) specifically recommends the use of role models within local communities to increase PA among females [28]. These recommendations also highlight the importance of ensuring sustainable engagement in PA through the development of school and community links. Interventions linking the school and wider community have shown to be effective in sustaining changes in PA levels among adolescents [11, 29], which ultimately requires effective partnership working within communities. That said, previous studies exploring perceptions of PA role models or school-based intervention designs are limited to adolescent populations and are manly US-based. There is a need therefore to design and implement preventative strategies to help evade the age-related decline in PA that is characteristic of adolescent girls.

To address the multiple level factors that can determine PA, a comprehensive approach is needed for the design of effective interventions. One approach is offered by Bronfenbrenner’s socio-ecological model [30, 31], which emphasises that an intervention should not only focus on the intrapersonal factors (for example, self-efficacy) but on the multiple layers which influence behaviour (for example, visibility of role models and PA community opportunities). To fully understand health behaviours and the contexts in which they occur, the socioecological model highlights the importance of gaining a full appreciation for interrelationships between individuals and the social, physical and policy environment, acknowledging that self-regulation is difficult to establish without wider social and institutional support [32]. Self-determination theory (SDT) has also been widely applied to school-aged children offering an effective framework for evaluating children’s PA behaviours at individual and contextual levels [33]. Highlighting the strength of the SDT in predicting PA, previous studies suggest that while the correlates of PA can be examined from a socioecological standpoint, SDT can be used to address psychosocial influences [34]. Approaches which are based on an empirical understanding of psychosocial factors underlying the target audience’s decision have been shown to increase their likeliness of being effective [35]. Theory integration enables the incorporation of multiple key constructs across theories in order to enhance our understanding of the underlying mechanisms of health behaviour change such as PA. Subsequently, Zhang and Solmon [33] suggest integrating SDT with the social ecological model to harness the strength of each theory.

Providing girls with active role models and links to existing community activities could positively influence girls PA levels before the anticipated decline in adolescence. This paper reports the results of formative research to inform the development and design of a school-based PA role model programme. The theoretical framework for the intervention design integrates SDT and the socio-ecological model. Gathering child, parent, teacher and wider stakeholder perspectives, we set out to: i) seek input on programme structure, content and delivery; ii) identify barriers and facilitators to intervention uptake and continued PA participation; and iii) use these data to subsequently inform the design of the first UK-based tailored PA intervention programme, CHARMING (CHoosing Active Role Models to INspire Girls), with the production of a logic model.

The section below is structured as follows: Firstly, results concerning the intervention design are summarised with Table 1 providing an overview of the main intervention components with exemplary quotes. These findings are mapped onto the intervention logic model (Fig. 1). Second, findings concerning intervention uptake and continued PA participation are presented according to each specific domain of the socio-ecological framework, with barriers and future strategies aligned (key themes and exemplary quotes denoted in Tables 2 and 3).

Reviews and meta-analyses have recently demonstrated larger effects among PA intervention designs which are; girls only, schools-based, multi-component and underpinned by theory [14, 41]. Calls for the inclusion of objective PA measures and greater involvement of preadolescent girls are also documented among reviews [42]. The literature identifies role models as a potential strategy to inspire young girls to become involved in, or maintain involvement in, PA and sport [27], with a recent longitudinal Australian study reporting that older girls were significantly more likely to be active if they had a role model [43]. Despite limited research exploring the use of role models to inspire PA participation and a lack of robust trials, policy recommendations strongly support the use of role models for tackling rising inactivity levels [42, 44]. Previous studies exploring perceptions of PA role models or school-based intervention designs are limited to adolescent populations. In an attempt to fill this understudied gap, the current study designed a school-based role model programme using a participative community approach with the target audience and stakeholders.

The first aim of this study was to gather views from preadolescent girls, parents, teachers and wider stakeholders in order to inform the design of a school-based, community-linked role model programme. Eliciting views on intervention structure, content and delivery, data revealed the importance of incorporating multiple viewpoints. Girls were clear that they wanted sessions to be fun and contain an element of competition. This finding is consistent with Jago et al. [45] who identified fun as a key motivator for PA among adolescent girls [46]. To ensure the programme is enjoyable, a key feature of the intervention structure is the provision of taster sessions, whereby a different activity is delivered each week. Girls believed that the programme would become boring with the same weekly activity and instead identified a range of appealing activities that they would like to attend. Providing girls with such ownership over the intervention design and ensuring sessions are fun and enjoyable, is consistent with multiple principles of SDT, including intrinsic motivation, relatedness needs, autonomy support and autonomous motivation [25]. In order to cater for the varying needs of a competitive element, each taster session should provide opportunities for competition and participation without an emphasis on winning. Adopted within previous intervention designs [47], this approach acknowledges that experiences of competence vary upon success or failure at challenging physical activities or as a function of feedback from the role model [48]. After-school was identified as the most suitable time to deliver taster sessions, particularly within the urban school. Teachers, parents and children suggested that this timing would enhance their ability to attend the sessions without encroaching on other commitments.

Consistent with other studies [47, 49, 50], data revealed that single-sex classes were deemed most likely to increase both intervention uptake and continued PA participation. The majority of girls reported that they would prefer single-sex sessions and one teacher expressed the perceived benefits, highlighting greater engagement of girls in previous single-sex school initiatives. Similarly, in other studies [49‐51], girls reported greater opportunities to develop skills and experience lowered self-consciousness and increased attention from session instructors in single-sex sessions. Camacho-Minano and colleagues [52] stress the importance of girl-only interventions not only to meet the unique needs of girls but also to avoid the typical PA contexts, such as praise for power, strength and aggressiveness [53], which tend to reinforce gender stereotypes and disadvantage girls.

Similarly, when discussing intervention delivery, this theme linked with the suggested requirement of sourcing female role models to lead sessions, with some girls reporting that they would feel more comfortable and have greater relatability about body concerns with female leads. Parents and teachers in the urban school also encouraged the use of only female role models, highlighting the complexities surrounding religious beliefs. Girls identified a range of characteristics that could inform the recruitment of role models, with interpersonal traits such as fun and friendly and behavioural traits such as helpful and kind deemed most important. Throughout discussions to identify role models, perceptions of personal connection were eminent, with findings directly underpinned by the relatedness principle of SDT.

Concerning the inclusion of health messages, teachers agreed that this component would form an important part of the intervention and fit appropriately into the school ethos. Both teachers identified the need to receive support from the research team in order to deliver messages during sessions. Discussions with policy makers highlighted the importance of ensuring messages were age-appropriate, clearly communicated and delivered in a contemporary manner.

The second aim was to examine the perceived barriers to intervention uptake and continued PA participation. Providing a framework for organising and representing key emergent themes, the socioecological model is an effective tool for conceptualising the many factors influencing PA behaviour [54]. Parents and teachers identified a number of barriers while discussions with stakeholders highlighted numerous strategies to help translate findings into practice. Within both schools, cost was identified as an overarching barrier, presenting across multiple domains of the socioecological model. Specifically, costs associated with session attendance (both at school and within the community), including the need for extra staffing and provision for siblings at community clubs, were perceived as significant barriers to both intervention uptake and continued PA participation. Cost is a commonly cited barrier in the literature concerning families from lower socioeconomic backgrounds [55, 56]. Other barriers identified included lack of transportation, timing of community clubs and family logistics, which are also consistent with previous research [37, 54]. Strategies to support children and their families may be required where access to transport for travel to and from community clubs is lacking [57].

Throughout the developmental process, we encountered considerable contextual complexities (e.g. different cultures, school locations, and single-sex staff). Stakeholder engagement was vital to ensure strategies addressed such complexities and that future implementation would reflect cultural contexts [19]. Within the urban school, cultural considerations commonly underpinned discussions and it was very clear that the use of female role models would positively influence the decisions of girls to participate, and most likely parents, to allow participation in a school-based role model programme. At the inter-personal level, parents play a vital role in supporting young children to be active through their attitudes, parenting practices and behaviours [58, 59]. In the current study, parents and teachers were ambivalent as to whether Muslim girls would be allowed to participate in such a PA programme as they considered it to be outside of the norm. Previous studies have detailed how Muslim women following their faith cannot engage in mixed gender sports and how some activities are deemed culturally unacceptable [60]. Research with religious scholars has also highlighted that while being active is supported by Islam, revealing parts of the body, often the case with sportswear, is deemed inappropriate for Muslim girls [61]. It is therefore important to assure parents that children will attend girl-only sessions and be permitted to wear cultural dress for the sessions. Furthermore, providing girls with a choice of activities which differ weekly will maximise the likelihood of participation in sessions. This data further highlighted the importance of tailoring at the local level and considerations for the cultural context of implementation [21]. Future work should examine Muslim girls’ participation in the intervention and or community clubs to gain greater insight in whether social, cultural or environmental factors present the greatest barrier to participation.

To enhance continued PA participation, the intervention should involve clear sign posting to community clubs and ensure school staff are aware of existing opportunities. Building partnerships between schools and community sport and recreation providers will help to facilitate links and future implementation [47]. One strategy is to formally link taster sessions with existing community sport and recreation providers; this will ensure consistency of role models, activities and available opportunities for girls. While exploring the barriers specific to this intervention component, the involvement of key stakeholders in the co-production process was a crucial mechanism for tailoring intervention strategies to the target group and intended delivery setting. The wide-ranging expertise and experience of stakeholders also identified existing initiatives for girls in the communities and provided insights into what has or has not worked well previously.

The third aim was to develop an intervention logic model outlining the design of the CHARMING study. During the co-production process, the research team reflected on the study data and input from all stakeholder engagement events. As emphasised by NICE guidelines [62], the intervention design needs to incorporate sound knowledge on the community needs and plans to build upon existing community resources all the while taking into account the local and national context. Creation of the intervention logic model provides a systematic overview of what the intervention will do, what the intended benefits are for the target population and will now be used to guide future development, implementation and evaluation.

All of the above findings will be used to specifically design and pilot the intervention for this population. Evident strengths of this study design are the triangulation of data (i.e. children, parents and teachers) and stakeholder consultation events. When designing a multi-component intervention that includes both a school and community element, partnership working is crucial and often depends upon mutual interests and investments [47]. While guiding intervention development, consultations with stakeholders aided the development of partnerships between organisations and enabled stakeholders to remain engaged with the project. This process has previously shown more results that are favourable in the long-term, with stakeholders being more likely to have an invested interest and make use of the generated evidence base [63, 64].

Some limitations to this study need to be acknowledged. In line with previous research showing the difficulties in engaging parents from low-socioeconomic areas [29], we encountered a low turnout of parents within the urban school. Prior to recruitment, the school staff had informed the research team of the difficulty they have historically had with engaging parents in school activities. Furthermore, all data were obtained from families living in two different areas within South Wales and no data were gathered from families of higher socioeconomic backgrounds, which limits our ability to generalise to other context or countries. The present study provides insights into how co-production with stakeholders can complement existing intervention development guidelines. While not surfacing in the present study, difficulties can arise when faced with competing priorities among stakeholders, this is a consideration for future intervention implementation and refinement particularly when sourcing community activities.

Based on the findings of the formative research presented in this study, the CHARMING intervention has been developed along with a logic model to guide subsequent testing and refinement. The intervention aims to increase PA levels and promote sustainable changes among preadolescent girls through role modelling and increasing awareness of opportunities for community PA. Future research must examine the feasibility of recruiting role models and tailoring design and implementation at the local level.