Health promotion for mild frailty based on behaviour change: Perceptions of older people and service providers

Abstract Mild frailty is common among older people, but it is potentially reversible with health promotion interventions. Behaviour change may be a key to preventing progression of frailty; however, we know little about what interventions work best and how a behaviour change approach would be perceived by this group. The aim of this study was to explore how mildly frail older people perceive health promotion based on behaviour change and what factors affect engagement with this approach. We conducted semi‐structured interviews with 16 older people with mild frailty who received a pilot home‐based behaviour change health promotion service, including a dyad of older person/family carer, and two service providers delivering the service in two diverse areas of South England. Interviews were audio‐recorded, transcribed and thematically analysed. The concept of goal setting was acceptable to most participants, though the process of goal setting needed time and consideration. Goals on maintaining independence, monitoring of progress and receiving feedback were reported to increase motivation. Physical/mental capability and knowledge/perception of own needs were main determinants of the type of goals chosen by participants as well as the approach used by the project workers. Older people with complex needs benefited from care coordination, with a combination of goal setting and elements of social, practical and emotional support in varying proportions. Mildly frail older people responded well to a behaviour change approach to promote health and well‐being. Further consideration is needed of the most effective strategies based on complexity of needs, and how to overcome barriers among people with cognitive impairment.


| INTRODUC TI ON
Frailty is an age-related condition characterised by loss of biological reserves across multiple organ systems, and vulnerability to physiological decompensation after a stressor event (Clegg, Young, Iliffe, Rikkert, & Rockwood, 2013). It is associated with an increased risk of hospitalisation, falls, moves to care homes and death (Ensrud et al., 2008;Rockwood et al., 2004). Frailty affects approximately 11% of people aged above 65 in developed countries, with approximately 41% considered pre-frail based on a systematic review of 15 studies from European countries, United States, Canada, Australia and Taiwan (Collard, Boter, Schoevers, & Oude Voshaar, 2012). Pre-frailty (mild frailty) is an intermediate state between being frail and robust (Fried et al., 2001). Older people with mild frailty can recover after a stressor event (e.g. minor injury, infection, new medication), but may increasingly rely on others for help with instrumental activities of daily living (IADLs) (Rockwood et al., 2005). Frailty is potentially reversible (Ng et al., 2015); complex interventions can reduce the risk of moving to care homes, hospital admission and falls in frail older people (Beswick et al., 2008). Although interventions targeted at mild frailty can potentially prevent frailty progression, there is insufficient evidence to recommend specific interventions .
A deficit model has traditionally been the main focus of available evidence informing health policies, the main disadvantage of which is it may be potentially disempowering for older people. As opposed to the deficit model, the asset model draws on the theory of salutogenesis (Antonovsky, 1996) to investigate key 'health assets' that support the creation of health rather than the prevention of disease (Morgan & Ziglio, 2007). According to Baltes' theory, successful ageing is the result of maintaining activities prioritised by the older person and their environment (selection), optimising the performance of these (optimisation) and compensating for limitations (compensation) (Baltes & Baltes, 1990).
Promoting health behaviour change depends on identifying strategies most likely to have positive impacts. According to the COM-B framework, all behaviour depends on three essential components: Capability, Opportunity and Motivation (Michie, Stralen, & West, 2011). Capability to perform the behaviour can be either 'physical' (having the physical skills, strength or stamina) or 'psychological' (having the knowledge, psychological skills, strength or stamina).
Opportunity is defined as "all the factors that lie outside the individual and make the behaviour possible or prompt it" (Michie et al., 2011, p. 4). Opportunity can be either 'physical' (what the environment allows or facilitates in terms of time, triggers, resources, locations, physical barriers, etc.) or 'social' (including interpersonal influences, social cues, cultural norms). Motivation is defined as "all those brain processes that energise and direct behaviour, not just goals and conscious decision-making" (Michie et al., 2011, p. 4). Motivation can be 'reflective' (involving self-conscious planning and evaluations) or 'automatic' (processing needs and desires, impulses and reflex responses) (Michie, Atkins, & West, 2014, pp. 59-60). Around this hub of capability, opportunity and motivation are situated nine intervention functions (broad category of means by which interventions may change behaviour): education, persuasion, incentivisation, coercion, training, enablement, modelling, environmental restructuring and restrictions (Michie et al., 2011).
Our systematic review described existing home-based health behaviour change interventions for older people with frailty or at risk of frailty, and examined links between their content and their potential to initiate behaviour change and improve outcomes . Drawing on recent developments in behavioural science, we classified intervention content according to the function(s) performed by the intervention and component behaviour change techniques (BCTs). BCTs are the irreducible intervention components that serve to perform one or more functions, such as setting goals or monitoring behaviour (Michie et al., 2013).
The BCTs 'provide instruction on how to perform a behaviour', 'add objects to the environment' and 'restructure the physical environment', and the intervention functions education and enablement showed most potential for improving physical function . This review revealed the lack of attention to behaviour change within intervention design, and to outcomes and mechanisms by which these complex interventions might work.
Despite several studies reporting older people's perceptions of health promotion interventions targeting specific problems, such as falls (Bunn, Dickinson, Barnett-Page, McInnes, & Horton, 2008;McMahon, Talley, & Wyman, 2011), smoking (Kerr, Watson, Tolson, Lough, & Brown, 2006), or reducing sedentary behaviour (Heseltine et al., 2015), very few studies explore how older people experience the delivery of BCTs in particular, within generic health promotion. A better understanding of older people's attitudes to and experiences What is known about the topic • Older people with mild frailty engage in activities to promote health and well-being, although ill health, social and environmental factors are potential barriers to maintaining independence.
• Behaviour change techniques are key components of health promotion interventions and may be helpful in preventing progression of frailty.

What this paper adds
• Findings from this qualitative study show that goal setting was perceived as a positive experience by the majority of mildly frail older people receiving a health promotion service based on behaviour change.
• Maintaining independence was a priority.
• Monitoring of progress and receiving feedback increased motivation.
of behaviour change would shed light on which strategies are considered most effective, and which techniques older people best engage with. A critical exploration of reasons for non-engagement can also contribute to improving the delivery of BCTs and hence maximise older people's benefits from health promotion interventions. This paper reports on a study that explored how health promotion based on behaviour change is perceived by both recipients and providers of a service, to understand how BCTs are received, which factors affect engagement with behaviour change strategies, and whether adjustments are needed for mildly frail older people.

| Study design and setting
This qualitative study was nested within a feasibility randomised controlled trial of a theory-and evidence-based home-based health promotion service for older people with mild frailty (HomeHealth) funded by Health Technology Assessment, National Institute for Health Research . As part of the mixed-methods evaluation of the service, we conducted qualitative interviews with intervention participants and service providers following the completion of the feasibility trial.

| The HomeHealth service
HomeHealth was a prototype of a new home-based health promotion service for older people with mild frailty, developed using a co-design approach and tailored to the individual's needs . The service was designed using the COM-B behaviour change theory (Michie et al., 2011) and was based on the principle of maintaining assets (Morgan & Ziglio, 2007). This was refined through one-to-one meetings with commissioners, managers and practitioners in urban and semi-rural areas and service development panels with frailer older people, health/socialcare and voluntary sector professionals, commissioners (funders), policy makers, academic experts and public representatives. The service was delivered in five to six appointments over a period of 6 months. The role of HomeHealth Project Worker was created for the purposes of the study. Criteria for the recruitment of project workers were excellent communication skills and previous experience of working with older people in community settings.
The two project workers received training in communication skills, physical activity and exercise, nutrition and behaviour change, with weekly supervision by an expert in older people and communication skills. The service was targeted at addressing four key domains: mobility, nutrition, socialising and psychological wellbeing, as well as other topics raised by participants. The project workers delivered tailored education, enablement, training and environmental restructuring as the main intervention functions.
The core set of BCTs used included goal setting, action planning and monitoring progress, maintenance of behaviours and developing habits. Goals were divided into: (a) an outcome goal (i.e. the overarching goal that the older person would like to achieve); (b) behavioural goal(s) (i.e. the specific action(s) agreed with the older person to achieve their outcome goal); and (c) SMART goals (detailed action plans that specify when, where, what and with whom the behavioural goal will be achieved). SMART goals are specific, measurable (recordable whether the goal was met or not), achievable (to build motivation), relevant (important to the person) and timely (achievable within a target time). The logic model of the HomeHealth intervention is published elsewhere, as have data relating to the practical and organisational aspects of delivering the HomeHealth service, including fidelity of delivery .

| Participant sampling, recruitment and procedure
We undertook face-to-face semi-structured qualitative interviews with older people who had received the HomeHealth service, and the two project workers who delivered the service. Participants were recruited from four different NHS general practices (two in urban and two in semi-rural areas) in London and Hertfordshire, UK. Eligibility criteria for entry into the HomeHealth trial were as follows: aged ≥65 registered with a participating general practice, scoring as mildly frail on the Clinical Frailty Scale (Rockwood et al., 2005)  Between 17 and 20 years 2 21 years and over 8 participants were screened for eligibility over the phone by a researcher using a set of questions based on the Fried frailty criteria (Fried et al., 2001), asking them about slowness, fatigue and if they needed assistance with ADLs. Baseline assessment was carried out at home, informed consent was sought and eligible participants entered the trial .
The mean age of participants recruited to the trial was 80 years at baseline. Baseline measurements showed that the sample was independent in ADLs based on Modified Barthel Index, but were demonstrating some signs of frailty with a lower grip strength and slower gait speed than the population average, and having an average of three to four long-term health conditions .
The additional criterion for entry into the present interview study was that participants must have received the HomeHealth service as part of the HomeHealth intervention trial. All 26 people who had received the service were invited to take part, and 16 consented to interview. Participants were interviewed individually, except one older person with dementia who was interviewed with their family carer.
Eighteen interviews were conducted (16 with older people, including a dyad of an older person and their carer and two project workers).
Participant characteristics are presented in Table 1.
All interviews followed the end of service delivery (August-December 2016). Topic guides were developed with input from the research team and refined iteratively as the study progressed, exploring participants' experiences of the service, motivation, identification of goals, how they found use of BCTs, and perceived impact of the service. Semi-structured interviews were broader in scope, but we have only extracted data relevant to our present research questions in this paper, as other data regarding fidelity have been reported elsewhere. Interviews with older people (conducted by CA or KK) took place in participants' homes with written informed consent. Participants were given a £20/$28 high street shopping voucher as thanks. Service provider interviews were conducted by JM in a private university office. Interviews were audio-recorded, transcribed verbatim and data were anonymised. The mean length of the interview was 49 min (range 23-87 min).

| Data analysis
We undertook thematic analysis of data using a six-stage process, including familiarisation with data, data coding, searching for themes, reviewing themes, defining and naming themes, and producing the report (Braun & Clarke, 2006). The multi-disciplinary analysis team (CA, KW, KK, RF, AL, RE) independently read transcripts and inductively identified a preliminary thematic framework to guide further coding.
This thematic framework was applied to a selection of transcripts by CA, further refined by the team and then applied by CA to all transcripts. COM-B and behaviour change theory were used to inform the analysis within a primarily inductive approach. The themes generated were then considered and interpreted by the team. NVivo software was used to manage the data (NVivo 11 Pro for Windows, QSR International). In this paper, we report on themes focusing on how older people with mild frailty experienced the behaviour change approach.

| Ethical approval
The study was approved by the NHS Camden and King's Cross Research Ethics Committee (ref. 14/LO/1698). All participants provided informed consent to participate in the study and anonymity was assured. Confidentiality was also assured subject to agreement that potential concerns about safeguarding/ill treatment of a vulnerable adult would be raised to appropriate authorities.

| FINDING S
We

| Information and perception of own needs
Knowledge about different topics played a role in how older people prioritised interventions. Provision of information by project workers was generally appreciated. It could form an important part of an overall comprehensive approach -for example in supporting an older person with caring responsibilities, education, problemsolving and help in coping with setbacks prompted her to make practical changes and helped her to cope better:

| Sources of motivation
Motivation to remain independent appeared to be an important determinant affecting engagement with a behaviour change approach.
Independence was represented as the ability to live without the assistance of other people. This was reported as the main desired outcome for many participants. Being able to still drive and get out was evidence of maintaining this ability, countering fears of losing independence and being seen as needing to move to long-term care:

| Goal setting
Goal setting was talked about positively by most participants:

| Monitoring and feedback on behaviour and outcomes
Most participants experienced monitoring progress towards goals and receiving feedback positively. As discussed above, this was a strong motivator but also triggered a proactive attitude towards dif- Positive feedback from the project workers appeared to increase participants' motivation and facilitate progress towards their goals.
Self-monitoring, for example by keeping a food diary or by recording walking activity, was reported helpful by some: …I felt that (project worker) was pleased that I was doing it. So it gave me more incentive to try and do a little bit more next time (OP6, Female, 72 years)

| Practical and emotional social support
Emotional support was important, especially to those living alone, those with mental health problems and those with complex needs (i.e. co-existence of different needs, related to mental health, social circumstances, physical health), who found that having someone taking an interest in them contributed to better recovery from illness.
The role of the project worker as a care coordinator was perceived as very helpful in addressing complicated problems including cognitive impairment, where practical social support was an essential part of the intervention: …I was present for one of those appointments, and I know that the follow-up from that was excellent. So More examples of participants' experiences with BCTs and the context in which they were used are presented in Table 2.

| D ISCUSS I ON
A home-based health promotion service for older people with mild frailty based on BCTs was overall well-received by recipients. The majority was able to identify a range of goals to work on, related to mobility, physical activity and transport, socialising, mental wellbeing, diet and finances. Most participants responded positively to goal setting, and only a few did not like the language or concept of setting goals. The process of goal setting could take time over several appointments. Most participants initially preferred the idea of setting goals to maintain their current independence/activities over the idea of improvement. Reflective listening and interaction with the same project worker over time enabled them to build a relationship of trust. Older people who had more complex needs reported benefiting from a combination of goal setting and elements of social, practical and emotional support (including helping co-ordinate their care) in varying proportions, with involvement of a carer in some instances.
In our development work, qualitative interviews with older people, carers and health professionals revealed that a health promotion service should cover a broad range of domains, and use mechanisms such as providing information and signposting, emotional and practical support, and boosting motivation . Although the COM-B model (Michie et al., 2011), which underpinned the design of HomeHealth, has been increasingly used for behaviour change health promotion across different disciplines, few studies explore older people's experiences with this approach, and it is the first to focus on people who are mildly frail. In our study, capability clearly determined older people's choice of goals. Inherent capability acted as a substrate for external opportunity, both in the sense of maintaining physical function and promoting health, as well as dealing with problems.
Although the conceptual framework of the development of the service was asset-based (Morgan & Ziglio, 2007), the choice of goals was probably not completely free of a deficit-oriented mentality, especially in older people whose physical health or cognitive functions had deteriorated recently. Opportunity was reported to increase as a result of receiving the service, through change of the physical environment, increasing social cues or combination Social support (emotional) Mental well-being "She was encouraging, she was telling me that I was doing well, improving. It's very nice to be encouraged, but I was -actually when she was with me, I was well because she is so nice. And I was enjoying her visits." (OP10, Female, 87 years) Social support (practical and emotional) Mental well-being "Well, first of all, she said about I shouldn't feel this guilt that I did, which, I mean when somebody tells you that, it does make you feel better. But she sort of encouraged me to, first of all, see a counsellor, which I did. Then she encouraged me to join things. And she gave me the information like the U3A, which I did join. So, that is how she helped me." (OP12, Female, 73 years) also been reported in younger populations (Sutcliffe et al., 2017).
The origin of motivation merits further exploration. Although the desire to remain independent was a commonly reported reason for entering the feasibility trial, most people emphasised the importance of the project worker acting as an external motivator, whose presence was key in introducing new ideas and creating a supportive framework to bring about change. Psychosocial support was also perceived as an important element of the intervention in a study exploring older people's experiences of a community matron primary care service (Williams, Smith, Chapman, & Oliver, 2011). Similarly, the importance of taking an interest in the older person was highlighted in a Swedish qualitative study about experiences of pre-frail very old people who received preventive home visits to identify unmet needs and provide local service information (Dahlin-Ivanoff et al., 2010). Interestingly, that study reported some people thought that such interventions were not for them because they were too ill or felt too old with nothing to anticipate (Behm, Ivanoff, & Zidén, 2013).
The level of need, defined by health and socioeconomic status, appears to be an important determinant of older people's engagement with health promotion. People who are 'too fit' may not engage with an intervention because they do not feel the need for doing so, whereas those with complex needs may struggle to engage because they face numerous barriers, leading to low levels of aspiration. In our study, older people's own subjective individual judgement about their needs was an important factor influencing their attitudes towards information provided by the project workers. Other research has shown that health maintenance-related goals are the most common, whereas people with better health resources are more likely to report goals related to leisure-time, social and physical activities, and those with poor social resources are at risk of having no personal goals (Saajanaho et al., 2016). These findings are in keeping with the theory of Ziegelmann and Knoll (2015) who distinguished health behaviours in two types: 'proximal', that is a core set of behaviours directly linked to physiological processes or producing straightforward health benefits, and 'distal', that is more complex activities indirectly linked to health-related outcomes via different pathways.
The hierarchy of goals is therefore inevitably shaped by the older person's perceptions of health and unmet needs. The reluctance of some participants in our study to identify themselves as being at risk of malnutrition can be explained by a different prioritisation of needs and desired outcomes from participation in a health promotion programme.

| Strengths and limitations
The main strength of this qualitative study is the novelty of the findings. We found that older people with mild frailty can engage well with behaviour change interventions, providing they are tailored to their physical health needs, and delivered over time by a support worker with good communication skills. We interviewed both recipients and providers of the HomeHealth service, which gave complementary insights into factors affecting engagement. However, the data are drawn from a small sample participating in the feasibility trial, who as volunteers are likely to be more motivated. Additionally, not all agreed to be interviewed, so other views regarding behaviour change may be missing, for example half of the interviewees had a high level of education, which is associated with having more personal goals (Lawton, Moss, Winter, & Hoffman, 2002). Moreover, given the nature of the sampling, and having a small pool of participants in a feasibility study, it was not possible to use saturation as a method to ascertain completion of data collection. Only two participants were given a diagnosis of cognitive impairment during the delivery of the service, therefore findings for this group need to be interpreted with caution as the full range of views for this population is unlikely to have been explored.

| Implications for practice and research
Results from this qualitative study are promising regarding the implementation of BCTs to promote health and well-being in mildly frail older people. Capability to undertake change is a key aspect, and older people with mild frailty appear to benefit with tailored support from a service to address this. Further research is now needed to determine if this approach is both clinically and cost-effective, before widespread implementation into routine care.

| CON CLUS ION
The majority of older people with mild frailty responded positively to a behaviour change approach delivered in the context of a home-based health promotion service. Goal setting, monitoring of behaviour and feedback were perceived to increase motivation.
Challenges of using goal setting in people with complex needs, including those with cognitive impairment, need to be accounted for when designing health promotion services. Practical and emotional social support to maximise capability should be included to promote health and well-being in this group of older people who are becoming frailer.

CO N FLI C T O F I NTE R E S T
We have no conflict of interest to declare.