Introduction
Physical activity is associated with reduced risks of obesity, cardiovascular diseases, type 2 diabetes, cancer and mortality [
1]. In the United Kingdom, national physical activity guidelines recommend that adults should be active every day with at least 150 min of moderate or vigorous activity per week [
2]. A target of 5 days per week with at least 30 min of activity is widely recommended [
2]. However, according to the World Health Organisation, one in four adults worldwide are insufficiently physically active [
1]. In England, 25.2% of adults were
‘physically inactive’ in 2017 with less than 30 min of physical activity per week [
3]. Women [
4‐
6], older people [
5,
7], non-white individuals [
6,
7] and those living in socioeconomically deprived neighbourhoods [
8] are more likely to be inactive.
Exercise referral schemes (ERS) represent a form of social prescribing that have been first set up around 1990 in the United Kingdom and is growing rapidly since then [
9]. ERS have been used in the context of health care services in attempts to increase physical activity in key target groups such as people with an elevated body mass index or people with specific conditions, for example depression or stroke [
10]. Previous research on ERS found differences based on sociodemographic factors between individuals who are more or less likely to be referred, take up the referral, adhere and complete the scheme and achieve changes as a result [
5,
11,
12]. For instance, older adults were more likely to take up the referrals and complete ERS compared to younger age groups [
5,
12]. Other differences were observed in relation to existing health conditions, where people with pulmonary diseases were less likely to complete ERS than people with cardiovascular diseases [
12]. Favourable aspect found in ERS research findings are the presence and importance of social support for participants [
13,
14]. Other factors were reported as barriers to whether participants take up the referral and complete ERS, including the timing of sessions, cost and location, awareness of the project, competing demands and intimidating/ unappealing/ uninviting gym environments [
14,
15]. Facilitators may include: internal motivation, support from providers, financial incentives, supportive environments and modified eating habits [
14,
16]. Most of previous research has been through randomised controlled trials or local evaluations. There is a lack of qualitative research into ERS that could help identify factors that may have affected engagement and outcomes of such schemes. Previous evidence has mainly focused on specific population experiences (e.g. people with disabilities), and only from the participants’ perspective [
14]. Combining participants’ perspectives with the scheme providers, including referrers, to identify a range of factors associated with referring participants, scheme implementation, attendance and adherence to the scheme. This could inform future developments of the same project or for the project’s use in another setting or population.
This study addresses the need to develop, implement and evaluate population-level interventions to increase physical activity with a focus on increasing engagement in exercise among individuals that are presently physically inactive. The present study reports a novel community-based ERS that was open to physically inactive residents in five socioeconomically deprived electoral wards in an inner London borough. The scheme, known as Active Lifestyles for All (ALFA), was one of 16 projects funded through Sport England’s Get Healthy Get Active scheme. This qualitative study aimed to explore the experiences of participants and staff involved in an ERS and to identify barriers and facilitators to the participation and the implementation of such projects.
Methods
ALFA project description
The ALFA project was a three-year community-led intervention in one London Borough that was managed by a social enterprise with expertise in sport education. The ALFA project was targeted at residents in five socioeconomically deprived electoral wards. All activities took place at the social enterprise’s new ‘healthy lifestyle centre’ in a regenerated playing field in the neighbourhood, which included exercise facilities, a café and space for social interaction. The location was in a quieter residential area a distance from main roads. It aimed to encourage physically inactive people aged 14 years and above to become and stay active. Potential participants were screened for eligibility using the Single Item Questionnaire for Physical Activity and were identified as inactive (i.e. completed 30 min or less of moderate activity a week). The ALFA project recruited a total of 1416 participants (including 1216 women and 200 men) between September 2015 and June 2018. The largest number of participants (618, 44%) were self-referred and the remaining participants were referred from the National Health Services including GPs, pharmacies and mental health services, public health services in the Borough, and community organisations for older people and people who are overweight and obese. There were 212 (15%) aged 65 years and older and 15 (1%) aged less than 18 years. The distribution of participants by ethnic group generally represented the demographic make-up of the local area, with approximately equal numbers of ‘white’ (476, 34%) and ‘black Caribbean and black African’ (485, 34%) ethnicity and smaller representations of Asian, mixed, other and not stated ethnic groups. Deprivation was evaluated using the Indices of Multiple deprivation for 2015, based on participants’ residential Lower Super Output Area, there were 486 (34%) in the lowest 5th of deprivation and 496 (35%) in the second lowest 5th of deprivation. Sport activity was recorded at baseline for 1265 (89%) participants using the International Physical Activity Questionnaire (IPAQ) questionnaire.
The project involved trained health mentors working with physically inactive residents to develop individually tailored exercise and health plans for becoming active through a choice of a range of sport activities. Mentors met with their project participants initially to discuss goals and take baseline physical activity assessments. Sports activities varied to suit a range of people of different physical abilities including yoga, fitness trampoline classes and Zumba with the option of a seated version for those less able to stand. The mentors’ role included supporting participants to engage in their selected sport activities for at least 30 min (i.e. one session) per week for 13 weeks free of charge. Subsequently, sessions were offered at a reduced rate. The project had no explicit theoretical framework and the fidelity of implementation was not explicitly addressed. There were 1110 (88%) who reported sport participation on no days per week; 125 (10%) participated on 1 day per week; and 30 (2%) participated on two or more days per week. Sport activity was recorded at 3 months for 1013 (72%) participants. There were 661 (65%) who reported sport participation on no days per week; 267 (26%) participated on 1 day per week; and 85 (8%) participated on two or more days per week.
Evaluation design
This study employed qualitative methodology, applying framework analysis to data collected from semi-structured interviews with stakeholders involved in the project. The study was approved by Kings College London Ethics Committee (Ref no. LRS14-15-1659).
Participant recruitment
There were two distinct groups of respondents namely: project participants and staff. Project participants were individuals that had been recruited into the ALFA project and had taken part or were currently taking part in their 13 activity sessions. They were recruited via the project manager who acted as a gatekeeper informing people of the study for the research team to then contact. The staff group were further distinguished into referrers, mentors and other staff. Referrers worked for external organisations that referred individuals to the ALFA project and included GP’s and community workers. Mentors on the ALFA project were staff members who were allocated to supervise a group of project participants. Other staff respondents in the scheme included those with management and consultation roles for the ALFA project. The project participants and staff were sampled for the interviews by the ALFA project manager using convenience and subsequently snowball sampling. Recruitment was planned to continue until no new themes arose and data saturation has been reached.
Interview schedule
The development of the interview schedule (Additional file
1) drew on the Theoretical Domains Framework (TDF) [
16]. The TDF is an integrated framework that aims to provide a comprehensive, theory-informed approach to identifying influences on behaviour, the TDF has been used in a range of studies [
17]. In the present study, the TDF was employed with the aim of enabling the identification of factors that participants may not otherwise report [
18]. For this study the most recent framework was used which included 14 domains, covering a broad range of individual and organizational theories [
19]. By using this framework, it limited the possibility of missing important factors affecting the participation in the project. The topic guide was discussed and edited by three experts in the field with experience in using the TDF in qualitative research.
Data collection
Interviews were conducted face-to-face at either the ALFA project centre, or at the staff member’s workplace. Interviews were conducted face-to-face, with the exception of two interviews conducted over the phone. Information sheets were sent to individuals prior to the interview and written consent was taken at the beginning of each interview. Verbal consent was taken for the telephone interviews. Interviews were conducted by one researcher (AGP) between October 2017 and February 2018. The interviews lasted from 10 to 55 min.
Data analysis
The interviews were audio-recorded and transcribed verbatim. The transcripts from the interviews were coded and managed using NVivo software [
20] and the coded data were managed and classified using Framework Analysis [
21]. The Framework Analysis process follows seven steps: transcription, familiarization, coding, developing and then applying an analytical framework, charting data and data interpretation [
21]. The interview transcripts were coded initially and refined by the first author (AGP) and checked by another author (SA) for consistency and a third author intended to resolve any remaining disagreement. Face-to-face meetings were conducted to discuss codes, and amendments were made where appropriate until full agreement was reached. The coded data were then examined in relation to the TDF and tables were produced for the development of the analytical framework. Data interpretation and the creation of themes and sub-themes was initially completed by the first author (AGP) and then checked, discussed and refined with the research team.
Discussion
This study aimed to evaluate factors that enabled or impeded participating in a community-based physical activity referral project using stakeholder perspectives. The study found that project participants valued the social aspects of project, including the way the project differed from ‘regular’ gyms in terms of cost, clientele and support provided. These factors acted as motivators to commit to the project and participants also felt a responsibility to complete the project because of the financial incentive and so as not to disappoint their mentors. At an individual level, personal circumstances influenced how they were able to fit the classes in their lives. Health conditions of the participants were acknowledged either as a barrier or facilitator by staff and the project participants. Changes as a result of the project included an increase in participants’ confidence to exercise and changes in their overall lifestyle, including their families. At a wider level, access issues were viewed as important by staff and project participants; the location of the project centre had some drawbacks with its residential positioning. Awareness and promotion of the project was regarded as low, and GPs did not seem to engage in the referral process, therefore missing an important group.
The project’s social environment played a major role in facilitating participants’ engagement with physical activity. The project was perceived as a safe and comfortable environment due to it attracting people of all shapes and abilities. Many studies have suggested an association between the gym environment and feeling intimidated and uncomfortable [
14,
22]. Gyms were perceived to typically attract fit, slim and young people and thus an unwelcoming environment for an older and inactive population [
14]. The participation of like-minded companions in group physical activity facilitated the enjoyment and integration to physical activity schemes [
14]. Providing a variety of classes that is suitable for participants’ differing needs and health conditions motivated the continuation in physical activity, as reported in the current study, and previous evidence [
14]. This study’s participants (both project participants and staff) have acknowledged the lack of staff skills in dealing with medical conditions that required special care. This could be the result of overambitious referral strategy and lack of communication between the referrers and the mentors. Therefore, such programmes need to be accompanied by skilled trainers with an experience of dealing with different medical conditions.
The method of recruitment may have played a role on the project participants’ experience. Self-referred participants appeared to be more motivated and engaged with the project. Project staff thought that participants who were referred by their GP were less committed and enthusiastic about the project. It has been previously suggested that the impact of health assessment on behaviour change may vary according to the mode of recruitment, where self-referred participants being more responsive and more likely to report health behaviour change [
22]. Self-referred participants might be more prepared to act and motivated to change than actively recruited participants. However, it might be unwise for future projects to rely only on self-referrals because then needier groups, with greater barriers to participation, might not be included.
Social support appears to be an important facilitator in engaging and motivating participants in the project. This came in the form of participants’ relationship with their mentors and fellow participants. The sense of care and continued follow-up was an essential factor to participant engagement, as reported previously [
14]. Participants relationship with their mentors created a sense of responsibility and loyalty to continue with the project. Although this factor facilitated the continuation of the project, the impact on the long-term participation in physical activity is unclear. The use of technology-based interventions to facilitate behaviour change should be considered to increase people’s access to and continuation with physical activity interventions. However, technology-based interventions should be of interest to the target population. Evidence has suggested that an easy to use website providing step-by-step videos and interactive feedback is a facilitator to exercise among middle aged individuals [
23]. These interventions could also be of benefit in providing professional support beyond the project, as people reported concerns over the lack of on-going support and it was seen as a barrier to continuing to exercise [
14].
Financial costs and physical inaccessibility are well established barriers to physical activity [
14,
23]. Providing free or subsidised exercise sessions incentivised the participation in the project. It has been suggested that targeting financial incentives to high-risk population may result in greater behaviour change success [
24]. However, physical activity is a habitual behaviour and the use of financial incentives may assist in initiating the behaviour but may not necessarily result in long term behaviour change. Although it has been suggested that financial incentives are associated with longer-term smoking cessation [
25], it is not clear whether changes in physical activity will be maintained. Lack of transportation and inconvenient timing of classes also contributed to the difficulty to access physical activity interventions [
14,
23]. Previous evidence has also suggested that neighbourhood safety and weather conditions are key factors contributing to inaccessibility [
14,
23]. Offering behaviour change programmes in multiple accessible locations could improve people’s engagement with physical activity. Furthermore, providing older adults and individuals with disabilities with appropriate transport may increase their engagement with the programme.
The coordination between primary care professionals and ERS provide a great opportunity for those wanting to change. Primary care presents an important platform to initiating health behaviour change due to its longitudinal nature. However, according to this study’s participants, GPs were not fully committed to referring eligible patients to the project. A recent qualitative study suggested that although primary care professionals believed in the importance of referring patients to external lifestyle services, several obstacles were acknowledged that were faced by patients and healthcare professionals [
26]. These difficulties included long waiting lists, service discontinuation due to budget cuts, costs, keeping track of what is available and uncertainty of whether or not patients will take up the referral [
26,
27]. Continued follow-up of referrals made and improving communications between general practices and referral services is needed.
Participants in ERS in this study have reported making changes to their physical activity levels and overall lifestyles regardless of the reported barriers. These changes have also influenced changes to their families’ lifestyles. Participants in this study experienced positive effects on their physical and mental health as a result of increased physical activity levels. Previous evidence has suggested that ERS have the potential not only to increase adherence to physical activity [
28,
29], but may also enhance long-term sustainability [
30].
This study provides evidence of the potential benefits of ERS in improving physical activity levels among inactive populations. Future research should examine the use of interventions that provides continuous support and follow-up, including technology-based interventions. This is to ensure adherence to physical activity in the long-term. ERS should provide staff with skills required to deal with people of different health conditions and physical abilities. Improving the links and communication between ERS staff and healthcare professionals, including referrers, is essential to improve engagement to the project and overall outcomes.
Strengths and limitations
This study is among the first to present qualitative findings relating to participants’ and providers’ experience with an ERS project. A strength of the study is that it included 35 in-depth interviews from the three distinct groups in the project. This helped to provide a comprehensive picture of the project operations and the challenges and facilitators faced by those involved. Using the TDF may have led to study participants discussing factors that may not be otherwise discussed [
18]. The TDF enables looking beyond individual factors and considers environmental and social influences on behaviour. Study codes were checked by a second researcher who was independent from the data collection process. Through a process of consensus, the two researchers reached agreement. This procedure improved the transparency and credibility of this study interpretation [
31].
The findings of this study may not be completely transferable beyond the current context. Our findings may be specific to this project and its setup and may not apply to other ERSs particularly those aimed at individuals with certain conditions e.g. diabetes. Due to data confidentiality, the project manager contacted and subsequently recruited participants for the interviews. This may have introduced bias and limited the representativeness of the sample from the project general population. Project participants who were not able to participate in the study may have had different experiences. We would also have liked to recruit more people in the study that had dropped out of the study at different points and more men participants for the study to establish potential barriers to entry, engaging and continuation in the project. Face to face interviews can produce social desirability bias [
32], especially when discussing health behaviour change and implementation behaviour. The interviewer, however, was external to the project and other related organisations, with no conflicting roles or affiliations, which is believed to help in accessing more private accounts. We acknowledge that collection of demographic data differed for staff and participants. For example, ethnicity and deprivation level were not collected for staff members. Finally, theoretical saturation was not known to have been reached due to a relatively small sample of project staff.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.