Article Text

Original research
Factors influencing adults who participate in a physical activity coaching intervention: a theoretically informed qualitative study
  1. Stephen Barrett1,2,
  2. Stephen Begg2,
  3. Paul O'Halloran2,
  4. Jeff Breckon3,
  5. Kane Rodda4,
  6. Gabrielle Barrett1,
  7. Michael Kingsley2,5
  1. 1Health Promotion Department, Bendigo Health Care Group, Bendigo, Victoria, Australia
  2. 2Holsworth Research Initiative, La Trobe University, Melbourne, Victoria, Australia
  3. 3Research and Innovation, Teesside University, Middlesbrough, North Yorkshire, UK
  4. 4Outpatient Rehabilitation Department, Bendigo Health Care Group, Bendigo, Victoria, Australia
  5. 5Department of Exercise Sciences, Newmarket, Auckland, New Zealand
  1. Correspondence to Professor Michael Kingsley; M.Kingsley{at}; michael.kingsley{at}


Objective Behaviour change interventions targeting changes in physical activity (PA) can benefit by examining the underlying mechanisms that promote change. This study explored the use of the Capability, Opportunity, Motivation and Behaviour (COM-B) model and the Theoretical Domains Framework (TDF) to code and contextualise the experiences of participants who completed a PA coaching intervention underpinned by motivational interviewing and cognitive–behavioural therapy.

Design Semistructured interviews were conducted with a purposive sample of participants.

Setting Interviews were conducted in a tertiary hospital in regional Victoria, Australia.

Participants Eighteen participants who completed a PA coaching intervention were interviewed. The participants were recruited into the coaching intervention because they were insufficiently physically active at the time of recruitment.

Results Thirteen (72%) participants were women and the average age of participants was 54 (±5) years. Four participant themes mapped directly onto five components of the COM-B model, and ten of the TDF domains. Increases in PA were influenced by changes in motivation and psychological capability. The autonomy-supportive PA coaching intervention helped to evoke participants’ own reasons (and motives) for change and influenced PA behaviours. Participants reflected on their own social and/or professional strengths, and used these skills to set appropriate PA goals and action plans. The structure of the PA coaching intervention provided clarity on session determinants and a framework from which to set an appropriate agenda. Relational components (eg, non-judgemental listening, collaboration) were continually highlighted as influential for change, and should be considered in future behaviour change intervention design.

Conclusions We demonstrate the beneficial effect of using theory-informed behaviour change techniques, and delivering them in a style that promotes autonomy and relatedness. The views of participants should be a key consideration in the design and implementation of PA coaching interventions

Trial registration number ACTRN12619000036112. Post-results analysis.

  • public health
  • preventive medicine
  • sports medicine

Data availability statement

Data are available upon reasonable request. The datasets used and/or analysed during the current study are available from the corresponding ethics organisation (Bendigo Health Care Group) via

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Strengths and limitations of this study

  • Semistructured interviews enabled in-depth exploration of the experiences of individuals who participated in a physical activity (PA) coaching intervention.

  • Including the perspectives of individuals who did not increase PA as well as those who did increase PA as a result of the intervention can be considered a strength of this study.

  • Another strength of this study is the application of the Capability, Opportunity, Motivation and Behaviour model and the Theoretical Domains Framework to explore the experiences of individuals that participated in a PA coaching intervention underpinned by motivational interviewing and cognitive–behavioural therapy.

  • The sample was exclusively non-admitted hospital patients that participated in a PA coaching intervention, and additional perspectives may provide a broader overview to inform intervention development.


Regular physical activity (PA) is positively associated with numerous health-related benefits and a marked reduction in risk for chronic disease.1 2 Although the importance of regular PA has been widely publicised, large numbers of adults do not undertake the recommended levels of PA.3 For example, only 50% of adults in the USA and 56% in Australia undertake the required amount of PA to be deemed sufficiently physically active.4 5 Behaviour change interventions have been increasingly used in an attempt to influence PA change; however, numerous systematic reviews and meta-analyses demonstrate that high proportions of participants revert to insufficient PA levels once the behaviour change intervention is ceased.6–8 The marked reductions or cessation of PA can nullify the health improvements gained from temporary PA increases.9 Consequently, there is a need to develop behaviour change interventions that strengthen the maintenance of PA over time.10

Rothman suggested that theory-informed (eg, social cognitive theory) interventions can be effective for promoting maintenance of behaviour change.11 Michie et al expanded on this more recently, proposing that behaviour change interventions need to address specific components of change, namely an individual’s capability, opportunity and motivation to change.12 The factors that influence initiation of behaviour change differ substantially to those that influence maintenance.12 13 Using the same theoretical constructs for behaviour change initiation and maintenance might not account for variations in capability, opportunity and motivation, many of which can be driven by intentions, past experience and environment.12 There is a need to explore the potential determinants of successful PA maintenance to assist in the development of interventions to produce lasting change.

Motivational interviewing (MI) was developed to elicit motivation for behaviour change from the individual.14 MI is an autonomy supportive intervention and seeks to empower clients to voice their own reasons and strategies for change.14 15 MI strategies for behaviour change maintenance are less specific, and are not emphasised as part of the four processes of MI (engaging, focusing, evoking and planning).14 As a result, the integration of action-orientated interventions such as cognitive–behavioural therapy (CBT) has been recommended.16 CBT involves assisting clients develop strategies and skills (eg, activity scheduling, successive approximation) to change behaviours.17 Instead of being passive recipients of CBT skills training, the integration of MI and CBT (MI-CBT) can ensure that clients’ have autonomy around the focus and direction of change, which might support maintained change.10 Integrated MI-CBT has demonstrated effectiveness for the maintenance of PA behaviour change across a number of studies18 19; however, participant’s experiences of the intervention are absent from the literature.

Qualitative approaches are beneficial to capture individuals’ in-depth perspectives of the phenomenon studied, in this case, their experience of the PA coaching intervention.20 This study examines the views of individuals who participated in an MI-CBT-based PA coaching intervention as part of a randomised controlled trial (RCT).19 Relative to control, the intervention group demonstrated significant changes in PA at post intervention (3 months) and these changes were maintained 9 months after the commencement of the intervention.19 At present, we do not understand what influence contextual factors (mode of delivery, behaviour change techniques (BCTs), therapeutic alliance) within the coaching intervention might have had on these findings. Thus, the aim of this current study was to qualitatively explore the experiences and perceptions of individuals who received the MI-CBT-based PA coaching intervention to identify determinants and facilitating factors that influenced PA behaviour change. These insights will provide a deeper understanding of their experiences, and might offer valuable information to assist health professionals to improve intervention effectiveness and uptake.


This study employed a qualitative design using an interpretive description approach,21 and adopted the Consolidated Criteria for Reporting Qualitative Research (online supplemental material 1).22 Semistructured interviews were conducted with a purposeful sample of adults who participated in a PA coaching intervention in the Healthy4U-2 (H4U-2) RCT to explore their experiences and perceptions of the intervention to identify factors that influenced PA behaviour change.

A detailed description of the PA coaching intervention (H4U-2) including the intervention schedule, theories and techniques is available in the published literature.19 In brief, in the H4U-2 study, 120 insufficiently active adults (aged 18–69) were recruited from an ambulatory hospital clinic and randomised to an intervention group that received an education session and PA coaching, or to a control group that received the education session only. The PA coaching intervention comprised integrated MI-CBT and was delivered in five 20 min sessions over 12 weeks via the telephone. The intervention used an MI framework and microskills (open-ended questions, affirmations, reflections and summaries) to underpin all sessions. The CBT skills training (eg, goal setting, problem-solving and coping strategies) was delivered using the MI framework. The intervention was delivered by a physiotherapist trained in MI-CBT through workshop attendances, and individual coaching from an experienced practicing psychologist.

The H4U-2 participants completed outcome measures at baseline, post intervention (3 months) and follow-up (9 months). This provided a 6-month non-intervention time to assess maintenance of behaviour change. At baseline, the mean age of the H4U-2 participants was 53±8 years and accelerometer-measured PA was 15±5 min/day of moderate-to-vigorous PA (MVPA). The intervention group increased MVPA at post intervention (23±10 min/day) and maintained this at 9-month follow-up (22±10 min/day). In contrast, the control group decreased MVPA at post intervention (13±6 min/day) and at follow-up (10±6 min/day).

Sampling and recruitment

All H4U-2 trial participants were asked to complete an evaluation form at the 9-month follow-up, which included a question about whether they would be willing to participate in a semistructured interview. Individuals who were in the intervention arm of the study and responded with ‘yes’ (n = 46) were considered as the sample eligible to participate in this study. A purposive sampling procedure was used. We aimed for a variation in the participants’ (1) change in PA (an increase, decrease or no change in PA between baseline and follow-up measured using accelerometers); (2) engagement with the behaviour change intervention, identified by the intervention provider; (3) gender and age to reflect the sample in the population; (4) geographic location (rural or regional); and (5) socioeconomic status, using postcodes as a proxy. A research assistant contacted the individuals to confirm their wish to participate. Permission was sought to give their contact details to the interviewers.

Twenty-five people were invited to participate. Two individuals declined; one reported no longer wishing to participate and the other individual cited health issues. The participant’s recruitment ceased when we met our variation sampling requirements and reached data saturation. Data saturation was considered to be reached when the analysis indicated that additional interviews were not providing new concepts and the data provided were sufficient to address the research aims.23 No new significant information was derived between the 17th and 18th interviews, indicating that data saturation was reached and interviewing was ceased.23

Interview process

Written informed consent was obtained from all participants at the start of the interviews. Face-to-face interviews were carried out in the health promotion department of the associated hospital between June and September 2020. The interviews were carried out by the first author. A semistructured interview guide was developed based on the existing literature24 25 and was used to facilitate the discussion (online supplemental material 2). The interview guide was piloted by interviewing three individuals who took part in an MI-CBT-based PA coaching intervention delivered by the health promotion department of the associated hospital; these three individuals did not participate in the H4U-2 study, but did receive the same 5×20 min session of PA coaching as the H4U-2 study participants, which was offered as part of standard health promotion practice in the associated hospital. Following this pilot, additional questions, probes and prompts were included to further explore individuals’ experiences in terms of engaging in the behaviour change intervention. The pilot interviews were not included in the final sample as the individuals were not enrolled into the study.

Data analysis

Interviews were audio recorded, transcribed verbatim and rendered anonymous. Participants were identified as participant 1, participant 2 and so forth. The transcripts were analysed using an interpretive description method.21 NVivo V.12 software (QSR International, Cambridge, Massachusetts, USA) was used to facilitate data analysis. The interpretive description approach requires emersion into the data to identify thematic patterns, and an inductive analysis to permit theorising about explanatory factors.21 We developed a draft coding frame to capture codes and emerging categories. The coding frame was trialled by authors (SBarrett and KR) who independently coded 20% of the transcripts. A revised version of the framework was developed and tested by the research team; this version was used to code all transcripts. The first author independently coded 18 transcripts, and 2 authors (KR and GB) independently coded 9 transcripts each. The level of agreement between independent coders was substantial. Disagreements were resolved through discussion among all coders.

Following this stage, the identified categories were mapped against the Theoretical Domains Framework (TDF)26 and Capability, Opportunity, Motivation and Behaviour (COM-B) model.12 The TDF and COM-B model provide the theoretical basis for understanding behaviour change. The TDF integrates 33 theories and 128 psychological constructs into 14 domains underpinned by psychological theory.26 The TDF domains include individual-level factors (knowledge, skills), social factors (social influences, social support), environment and resource factors (cost of resources to facilitate change).26 Taken together, the 14 domains prompt the consideration of a wide range of influences on behaviour change. The 14 validated domains are included in online supplemental material 3.

Michie et al identified three components that need to be present to influence behaviour (B): capability (C), opportunity (O) and motivation (M); together these components make up the COM-B model.12 Capability refers to having the knowledge and skills required to engage in a behaviour; it can be broken down into two components, psychological capability and physical capability. Opportunity refers to the external factors, which make undertaking a behaviour possible. Its two components are physical opportunity and social opportunity. Motivation refers to the internal processes which influence decision-making and behaviours. Its two components are reflective motivation and automatic motivation. The COM-B model is widely used to contextualise individual-level change and determine what needs to change for behaviour change interventions to be effective.12

We mapped the broad categories onto the TDF domains, as well as directly onto the six components of the COM-B model to identify factors that are likely to influence PA behaviour change and could be targeted in future interventions. Participants’ responses were analysed for descriptions of BCTs and skills used as part of the PA coaching intervention. The technical components of the interventions were mapped against Michie’s taxonomy of BCT,27 and MI processes, relational components and microskills.28 The mapping of draft themes and findings were discussed among all authors to investigate a broad perspective on thematic interpretations. Disagreements between the team were resolved through discussion. The vigour of the qualitative research was strengthened through the implementation of independent coding; triangulation of data, and the critical appraisal of developing themes.29

Patient and public involvement

The research was designed and conducted without patient or public involvement.


Eighteen people participated in semistructured interviews, where 13 (72%) were women and 5 (28%) were men. The average age of participants was 54 (±5) years, with participants ranging in age from 42 to 66 years. Table 1 provides details of the participants’ characteristics. The interviews ranged in duration from 26 to 45 min, with an average duration of 34 min.

Table 1

Profile characteristics of participants (N=18)

Perceptions of the PA coaching intervention were positive, including those participants who did not find the intervention beneficial for maintaining PA change. The structural components (ie, defined session times and parameters) of the PA coaching intervention provided the participants with a clear indication of what was involved, while the relational components (ie, MI spirit: collaboration, empathy, evocation, autonomy) provided a platform for the participants to engage and focus on their reasons for change. The structural and relational components highlighted by participants are detailed in table 2. The delivery of the intervention via phone was reported as favourable by the majority of the participants. The telephone delivery provided flexibility around participation in the intervention with 16 participants in gainful employment during the intervention period.

Table 2

Technical, relational and structural components of the physical activity coaching intervention

Four themes were identified that mapped across 5 of the COM-B components and 10 domains of the TDF framework (table 3). The themes included strength-based coaching, autonomy-supportive listening, reframing PA goals and self-regulation. As evidenced in table 3, the themes mapped to multiple COM-B components and TDF domains; as such the findings are presented below under the identified COM-B heading, and related TDF domain subheading. These findings are described in detail below and supported using illustrative quotes from participants. The participant’s gender, age and change in PA as a result of the PA coaching intervention are provided alongside each quote. The categories that informed these themes are presented in online supplemental material 4.

Table 3

Mapping of themes to TDF domains and COM-B model with the associated BCTs

Capability: psychological


Increased knowledge was not described as a key component that participants needed in order to increase PA. The participants repeatedly stated that they did not need to be told they would benefit from increasing PA.

Anyone who is unfit knows they need to get fit – simple. But just telling them won’t make a difference. (Female, 51, no change in PA)

The participants highlighted how the PA coaching intervention used their existing skills to facilitate changes in PA. This strength-based, person-centred approach did not seek to impart knowledge, rather it sought to evoke personal capabilities from the individual, and to build their autonomy in applying this knowledge towards initiating and sustaining changes in PA.

The coaching was good, and I think spent lots of time looking at things I was good at, and sort of, how to apply these to my exercise. But it forced me to have a good look at myself, and what I was good at and probably not too good at, and at the time it took me a while to be comfortable with both. (Male, 52, increased PA)


Although the individuals were recruited from a secondary care hospital clinic, physical capability was not discussed as a barrier to increasing PA. The participants did not indicate that they needed to be provided education on what exercise to do, or the skills required to do it. Instead, participants noted how in the past they felt that they did not have the ability to increase or maintain PA from a psychological or emotional perspective.

What was stopping me from being fit before…it was the mind. Not the body. I could physically do it, but not mentally. (Male, 50, increased PA)

Similar to the construct of knowledge, participants expressed how the intervention strategies sought to evoke from them examples of the skills that they possessed.

On paper being fit is easy right. You want to walk, get up and walk. And yet I sat at home on the couch feeling bad for myself for not exercising, even though I can walk. Me, a grown woman, house, kids, job. Can run them all and I wasn’t able get off the couch. So I needed a shift in perspective. (Female, 55, no change in PA)

Behavioural regulation

Many of the participants described making numerous attempts at increasing PA in the past, only to lapse back to being insufficiently physically active. The inability to maintain regular PA over repeated attempts had diminished their self-efficacy to be physically active. The MI-CBT intervention used in the H4U-2 study employed specific BCTs to influence behaviour regulation as early as session 3 (week 4 of 12) of the intervention. The process of exploring capability to maintain PA changes was new to many participants, as exemplified by the following quote:

I think we weren’t long into it, and I had started to do some exercise. And I was feeling good. He [intervention provider] started asking me about how I would manage to be fit if something happened, can’t remember exactly, say, my strength training class finished up or something. And I was thinking, shouldn’t he be telling me I was doing a good job. But you know, when you have to think about it, and explain yourself out loud it gets the wheels turning, and you have to think ‘how would I do it?’ Because I’ve slipped off the wagon before. (Female, 49, increased PA)

Opportunity: physical

Environmental context and resources

Physical opportunities were discussed alongside the shift in perspectives that many participants encountered, from PA being something that they had to do, to something that they wanted to do. The reframing of PA to something that was attainable and enjoyable resulted in individuals placing a higher value on PA; when PA was afforded a higher value, people made time for it irrespective of previously cited situational or environmental barriers.

I used to drive to work every day. Its 2.2 KMs, which I know now because of walking. And of course you know how bad parking can be, so I’d probably spend ages looking for a park. Now I walk every day, to and from work. If it rains I can bring brolly, or drive if it’s really bad. But I don’t think of driving anymore, I enjoy the walk and it’s a part of my day. (Female, 46, increased PA)

Opportunity: social

Social influences

Some of the participants enrolled in community exercise programmes, for example, walking groups and strength training programmes. The building of social links within groups was highlighted by some as an important factor for continuing attendance, though many others were not concerned with building social connections within PA groups.

A large part of what kept me going back to the group was the friendship I made with other people. I was only there for a few sessions and a bunch of them invited me to come along for coffee afterwards. That was always helpful in getting me along to sessions. (Female, 58, increased PA)

I did wonder if sometimes there was a bit too much chat about exercise groups and exercising with other people. If I’m going to exercise I’ll do it for myself, I don’t need to be going with someone for company. That’s not important for me. (Male, 54; decreased PA)

The social influence of partners and significant others was regularly discussed, primarily as it related to the provision of unsolicited, didactic advice. The participants were aware that they were not undertaking sufficient PA, they did not need this to be pointed out by their partners and significant others. This didactic style of support contrasts with the relational components of the PA coaching intervention such as autonomy-supportive listening and collaboration, components that participants found beneficial in influencing PA change.

There is always time for the right advice, and pointless advice is, well… pointless. (Female, 62, no change in PA)

My husband was with me on the appointment, and he thought it was great that the surgeon discussed me getting fit. He’s been on my back for a while about it. So, he is often asking when I’ll do more exercise, but I’ll tell you what, that makes me want to do even less in spite of him. Even though I knew I needed to do it. The fact that he was telling me to do more made me want to do less to show him. (Female, 48, increased PA)

Motivation: automatic


The repeated attempts to engage in regular PA resulted in a feeling of helplessness in many of the participants who enrolled in the PA coaching intervention. The sense of disappointment expressed by participants was compounded by the fact that the barriers to regular PA were not physical, but psychological and specifically emotional. The inability to maintain regular PA likely influenced the participants’ PA self-efficacy levels, and by extension their automatic motivation.

When you have a few cracks at it, and you keep ending up in the same place, it doesn’t feel good. You tell yourself, and maybe a friend ‘that’s it, I’m going to get myself in shape’. Then two months later they ask you how that’s going and you are ashamed that you haven’t done a thing. And that does nothing for the self-confidence. (Female, 46, increased PA)

I’m not afraid to say it, I needed the support. I mean, if I didn’t I wouldn’t have joined the coaching. I just wasn’t able to do it alone. My motivation was shot, I wasn’t, maybe, thinking clearly about it. Probably fed up and disappointed trying the same thing over and over and not going anywhere. (Male, 50, no change in PA)

Motivation: reflective

Intentions and goals

Many participants described how previous attempts at PA were driven by a desire to lose weight. When these attempts did not result in sustained PA levels or weight loss, they were sources of frustration. It was commonly reported that the intervention looked to reframe PA from a weight loss tool to something that might deliver general health benefits. This was followed with a change in goal setting for outcomes. Shifting goal setting away from weight loss metrics changed the overall intention of being active, and having flexible goals removed the notion of a binary outcome of meeting or not meeting goals.

For me, exercise was always about weight. Always. If I didn’t reach the target I failed. And generally, I didn’t meet the target. So that was no good. But now I don’t exercise for weight; I exercise for me. To make me feel good, and I am so much better for it. (Female, 55, increased PA)

Social/professional role and identity

The concept of social or professional role was commonly described by participants as a tool used within the PA coaching intervention to elicit from participants’ areas of their life that they felt they had achieved or maintained success in, and formulated part of the strengths-based approach to making PA changes. Many participants were in full-time employment with partners and dependent children. They commented how the PA intervention required them to reflect on the strengths they have and commonly use in these social and/or professional roles. This approach evokes personal or professional capabilities, and the participants were able to reflect on where a transference of skills to PA was possible.

So, they didn’t use these words, but I needed to think on the lines of ‘how can I run a team, which I think is reasonably successful, and I can’t get myself in shape’. Not those words, but there was some, maybe, probing maybe. And you know, I thought, ‘hey I am someone, and I can do this’. It wasn’t a kick in the bum, but it made me sit up a bit. (Female, 51, increased PA)

Beliefs about capability

Participants’ beliefs about their capabilities were closely linked to the TDF domains of intentions and goals. As an example, exercising to lose weight and failing to attain weight loss goals resulted in a negative mindset and low self-efficacy. By reviewing and revising behavioural goals and learning from previous attempts, the participants noted a shift in their perceptions on capability. When discussing their attempts to be active before the PA coaching intervention some of the foremost feelings that prevailed were those of frustration and disappointment. By reframing intentions and altering their goals participants expressed a greater degree of confidence in their ability to maintain regular PA.

Once I got going, then talking about reviewing the goals, and modifying, making harder, or easier as needed, [that] was all fine. The actual tasks right, the exercise itself, or the goal setting – they aren’t hard to do. It’s not overly complicated. But… but you need to be smart about it and sometimes we get ourselves into a right spot that we can’t see the timber from the forest. And the coaching can pull your head in a bit and give you perspective. (Female, 52, increased PA)


The aim of this study was to explore participants’ perceptions of the factors that influenced their behaviour change throughout the course of a PA coaching intervention delivered using an MI-CBT framework. The interviews identified a wide range of barriers that influenced participants’ capability, opportunity and motivation to undertake regular PA, as well as the key elements of the PA coaching intervention that addressed these barriers. These key elements identified by participants included the combination of relational factors, namely the MI spirit used to underpin the intervention, and the technical factors which were the CBT skills that were used. The PA coaching intervention was designed to ensure that MI and CBT were integrated together in all sessions, and the fidelity of delivery was measured.19 The participants highlighted the synergistic advantages of receiving both MI and CBT techniques to motivate and promote lasting PA change.

Changes in reflective motivation was one of the dominant components that influenced participants’ behaviour change throughout the course of the intervention. Participants attributed a major change in intentions towards PA because of the intervention, specifically the shifting of perspectives away from PA being something they had to do (eg, participants highlighted the erroneous assumption that PA was necessary to lose weight) to PA being something they valued.30 By reframing their intentions, the participants set goals appropriate to their needs and importantly, their sense of capability.30 This relates closely with the need for self-efficacy for behaviour change.31 32 Participants consistently noted that the strengths-orientated approach of the MI-CBT intervention helped build this self-belief. The evocation of personal capabilities is encouraged within MI14 15; participants stated that reflecting on their capabilities within their social and/or professional role highlighted their existing strengths. This demonstrated to individuals how to recognise the self-regulatory skills they already had, and how these skills could be transferred to the attainment of regular PA.33 Instead of providing expert advice, the integration of MI-CBT permitted the individuals to voice their intentions and goals, and to understand what skills were needed to achieve these goals.10 34 Positive perceptions in regard to one’s own abilities have been shown to increase the likelihood of longer-term PA behaviour change.35

In contrast to physical capability, which was rarely brought up in the interviews, the influence that psychological capability had on PA behaviour change was widely discussed by most participants. The participants in this study did not seek exercise prescription from the intervention, indeed they largely claimed to already know how to undertake their preferred exercise. The prevailing issue was the inter-relation between the lack of motivation and a decreased sense of psychological capability, resulting in them not exercising. When discussing how the intervention helped change behaviour, reflection on their personal capabilities and skills was a common technique mentioned by the participants. BCTs such as identification of self as role model have been demonstrated to influence change27; the use of MI techniques to evoke examples of personal capacity might have influenced their confidence in their own capabilities.28 MI has been shown to support psychological needs based on self-determination theory36 as well as enhance self-efficacy,32 and the integration of MI-CBT provides an autonomy-supportive framework for the delivery of BCTs.10

The interviews highlighted the importance of the relational components of the PA coaching intervention, namely the MI style or spirit (ie, collaboration, compassion, evocation and empathy) used to underpin the intervention delivery. For most participants, the intervention was viewed as a positive experience, with a strong emphasis on the value of being listened to.37 This reinforces the significance of person-centred interventions, and aligns with the importance of autonomy-supportive influences described within self-determination theory.36 Self-determination theory posits that the quality of the support influences motivation and can help build self-efficacy.36 The favourable experience of the intervention contrasts to the participants’ description of support offered by partners or significant others. This support was didactic in nature, and they felt they were being informed of what they should do, without being listened to. The provision of listening support is a fundamental MI technique, and the spirit of MI communicates compassion, acceptance and partnership.14 38 Using MI as the foundational platform provides a supportive environment in which to deliver non-judgemental understanding and empathy.38 These relational components are likely to result in an increased sense of autonomy and build reflective motivation to increase and maintain PA changes.10 28 38 39

Increased beliefs about capability and use of self-regulation strategies characterised participants who were successful in maintaining their PA, which included a 6-month non-intervention period from the end of the intervention to the final measurement. Perceptions of capability and motivation are some of the internal and external processes (cognitive, self-reflective and self-regulatory) that come into play in human psychosocial functioning.12 Indeed, behaviour or self-regulation has been shown to mediate PA behaviours.40 Some of the self-regulation strategies highlighted by participants included ‘relaxed’ goal setting and planning. The MI framework of the intervention encouraged the individuals to set goals appropriate for them, and work out and plan their own strategies to regulate their PA. This is consistent with the autonomy-supportive approach of MI by offering choice over goals and demonstrating to participants that there are different ways to achieve these goals.14 38 By empowering participants to set appropriate goals and demonstrating that it is the participant who decides what choices to make, the participants are likely to be more engaged in the process and the more demanding CBT elements of the intervention.41

Applied implications

From a policy perspective, the delivery of the intervention 1:1 over the telephone was found to be favourable for most participants. This permitted the development of an interpersonal relationship, which has been demonstrated to influence change,42 and at the same time the telephone delivery provided participants with a sense of physical space, which helped them relax. For health-service delivery, is it encouraging that many participants found the coaching via telephone effective as telephone delivery permits offering services to wide geographic regions and can be delivered at participants’ convenience.

Participants commonly noted that the provision of defined parameters around the intervention sessions was beneficial for instilling a degree of focus within the sessions. The participants enjoyed the autonomy of goal setting and planning; goal setting is not always associated with autonomy, and goals that are not self-endorsed are likely to inhibit motivation.43 Using MI to underpin the delivery of the PA coaching intervention likely contributed to the sense of empowerment the participants detailed they had in setting their own goals and agendas.30 Alongside this autonomy, the participants valued being provided with clarity on the scope of each intervention session as this provided a scaffold from which to establish their agenda. Providing clear descriptions for sessions, including scope, summaries and between session plans are key components for intervention fidelity for MI-CBT.44 Practitioners and researchers can be encouraged that the structural parameters provided from the fidelity framework were positively received by individuals, and potentially contributed to successful behaviour change.

From a practice perspective, a large proportion of the techniques identified by participants as being important for promoting PA change were classified as relational. The importance of relational components within interventions needs to be considered when promoting behaviour change.45 The techniques classified in well-established BCT taxonomies have centred on the content of interventions, and have not examined the interpersonal components of interventions.45 The PA coaching intervention in the H4U-2 study used established BCTs, but delivered them using an MI framework.19 The MI spirit is a style of interaction that promotes an interpersonal relationship; it represents the way that the intervention content is delivered.38 A number of authors have proposed that relational components of interventions are likely to interact with technical components to influence behaviour change.45 46 This is consistent with the argument put forward by Hilton and Johnston that it is important how behaviour change interventions are delivered, rather than exactly what is contained in the intervention.47 Integrating MI with CBT permits the combination of content and relational techniques to increase the effectiveness of the intervention.10 In this paper, we have examined participants’ experiences of the intervention and attempted to make the distinction between relational and content-based techniques found to influence behaviour change.

Strengths and limitations

By identifying BCT techniques and mapping successful intervention components to TDF domains, COM-B components and central tenets of self-determination theory, we have distilled some of the macrolevel what of behaviour change interventions down to the more microlevel of how. Using the TDF provides a deeper understanding of the barriers and enablers to PA for insufficiently physically active ambulatory care patients.26 Mapping the findings into the COM-B model and highlighting specific BCTs is a significant strength of this study due to the integration of theoretically derived domains and structural and relational BCTs. Together, they demonstrate the theory-informed use of MI-CBT as an evidence-based intervention to increase and maintain PA. The design of the PA coaching intervention was based on determinants of PA change12 14 19; gaining perspectives from individuals who participated in the intervention provides further evidence to assist in developing effective interventions in the future.

There were some potential limitations in this study. Some difficulty arose in the categorisation of TDF themes and associated BCTs due to a degree of ambiguity in the definitions of the theoretical domains. Where this arose, the categorisation was determined through consensus via discussion within the research team. The sampling frame for the study was another potential limitation as all participants were recruited through an ambulatory hospital clinic in one hospital setting. The recruitment from one setting only potentially restricted the diversity in participants, in particular diversity in ethnicity, and limits the generalisability of the findings to broader populations. The participants who agreed to take part in this qualitative study may have been motivated and willing to talk about PA. The H4U-2 trial participants were asked if they were willing to participate in this study and we purposely recruited participants who did not increase PA as a result of the intervention to provide a broad sample. We made a conscious effort to recruit male and female participants, to provide understanding into the experiences of both genders.


This study provides an understanding of how participants perceived a PA coaching intervention and identified some of the behavioural factors that enabled or inhibited PA and the components that influenced their PA behaviour change. Using the TDF and COM-B model provides a theoretical basis for understanding behaviour factors in specific contexts, providing an indication of what is required to change. Identifying content and relational BCTs provides an overview of how to deliver autonomy-supportive interventions to support self-regulation of PA behaviour and build self-efficacy to maintain change. The findings from this study are valuable from theoretical, applied, training and commissioning perspectives because the BCTs, and the structural and relational components of the intervention that influenced behaviour change were identified first-hand by participants involved in the study. Its findings can be used to influence future intervention design, delivery and its monitoring and evaluation.

Data availability statement

Data are available upon reasonable request. The datasets used and/or analysed during the current study are available from the corresponding ethics organisation (Bendigo Health Care Group) via

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by Ethics Committees, Bendigo Health Care Group. Reference number LNR/18/BHCG/44121. La Trobe University College of Science Health and Engineering Human Ethics. Participants gave informed consent to participate in the study before taking part.


Supplementary materials


  • Twitter @barrettstephen1, @jbreckon

  • Contributors SBarrett, MK, SBegg and PO’H conceived the project and assisted with the protocol design. SBarrett collected and analysed the data with support from KR and GB. SBarrett wrote the first draft of the manuscript. SBarrett, SBegg, PO’H, JB, KR, GB and MK critically reviewed the manuscript and provided detailed feedback. All authors read, edited and approved the final manuscript as submitted. SBarrett, SBegg and MK are the guarantors of the article.

  • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. We acknowledge the support of the Bendigo Tertiary Education Anniversary Foundation and Holsworth Research Initiative for MK’s research.

  • Competing interests None declared.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.