Towards co‐designing active ageing strategies: A qualitative study to develop a meaningful physical activity typology for later life

Abstract Background Physical activity levels decline in later life despite the known benefits for physical, cognitive and mental health. Older people find it difficult to meet activity targets; therefore, more realistic and meaningful strategies are needed. We aimed to develop a typology of older people's motivations and lifelong habits of being active as a starting point to co‐designing active ageing strategies in a workshop. Methods We conducted semi‐structured interviews with 27 participants aged 65‐80 in Norfolk, UK, and participant observation with 17 of them. At a workshop with 13 study participants and 6 government and civil society representatives, we invited reflections on preliminary findings. Results Three types were developed. “Exercisers” had engaged in sport and exercise throughout their life but experienced physical ill health and limitations as barriers. “Out‐and‐about‐ers” pursued social engagement and a variety of interests but experienced biographical disruption through retirement and loss of companions that limited social activities in later life. A final type characterized people who preferred “sedentary/solitary” activities. A workshop elicited suggestions for new strategies relating to these types that addressed people's specific motivations. An example was to combine social engagement and physical activity in “dog‐parent”‐walking schemes to link people through shared responsibility for a dog. Conclusions We suggest that these potential strategies map more closely onto the everyday life‐worlds in which public health might seek to intervene than common physical activity interventions. Most notably, this means a more differentiated understanding of barriers, and acknowledging that intellectual, social or solitary pursuits can include incidental physical activity.


| INTRODUC TI ON
Our populations are ageing, and leading active lives is considered a "best buy" for preventing chronic conditions such as diabetes, heart disease and cancers, cognitive decline and dementia as well as social isolation and mental ill health. [1][2][3][4][5] Many older people, however, find it difficult to meet current physical activity guidelines. 6 Systematic reviews suggest that, to date, interventions promoting physical activity have typically produced only small or short-lived behaviour change. 7,8 Yet these reviews tell us little as to why success has been so limited. Public health intervention strategies tend to target either "at-risk" individuals, often focused on structured activities (eg, gym referrals), or "the population" at large, for example through health promotion campaigns (eg, to encourage walking) or through providing more supportive physical environments (eg, cycle lanes). 9 An underlying assumption of most physical activity interventions seems to be that people who are "insufficiently active" might need to learn about the benefits of being active or might require more appealing or accessible opportunities for sports and exercise.
What seems to be missing from the development of many intervention strategies is a greater understanding of people's dispositions, aspirations and life-worlds that underlie their physical activity patterns. People who are considered not to meet recommended activity levels cannot be assumed to share similar preferences or constraints, and this is where current interventions might fall short. While quantitative physical activity studies generally classify the population into those meeting or failing to meet particular physical activity targets, and some studies attempt to characterize participants into activity profiles (eg, by ability, physical activity domain or demographic characteristics), [10][11][12][13] these are usually based on people's current (reported or measured) activity level. People subsumed as "inactive", however, might be so for a variety of reasons, which could not be addressed through the same strategy. Older people might experience a variety of barriers from physical limitations to social isolation and neither necessarily related to their attitudes towards physical activity. Developing a more in-depth insight into people's experiences would entail moving beyond a broad-brush, generic approach to individuals, communities or populations to actively involve people in research that aims to understand how health behaviour is shaped by their biographies and deeply embedded in their social and physical environments. 14 This might also help to interrogate the inherent assumptions underlying common intervention strategies and could inform future strategies that map more closely onto the everyday life-worlds in which 1 might seek to intervene.
The overall aim of this study was to better understand the varied experiences of declining physical activity in later and to develop a differentiated approach to addressing barriers in more realistic and meaningful ways. First, we undertook qualitative research to explore lifelong practices and aspirations of active ageing, developed a motivation-based typology for being active and described respective barriers to staying active into older life. In a second part to this study, we held a workshop in which we invited participants to contribute to the interpretation of their data. The workshop was conceived of as an exercise in collaborative analysis and reflection to jointly discuss new directions for intervention strategies to promote active living.

| Participants
This qualitative study was embedded in a larger epidemiological project on physical activity in later life and its health outcomes, which in turn made use of data from the European Prospective Investigation into Cancer and Nutrition (EPIC)-Norfolk study. 15 We purposefully sampled a diverse, information-rich sample of 27 women and men, aged 65-80 years, classified as belonging to a professional or manual occupational class, living alone or cohabiting, in urban or rural neighbourhoods in Norfolk, UK, and by physical activity level. The EPIC-Norfolk study coordination team initially mailed 32 potential participants, and 22 agreed to take part; another 8 were invited in a second round of recruitment to fill under-represented categories (inactive, manual occupational class and oldest age group), of whom 5 responded positively. 16 For purposive sampling, we had access to EPIC participants' demographic data and their latest accelerometer measures of physical activity (see Table 1) which we used to sample participants belonging to a more (but not the most) active group or a more (but not the most) inactive group. This should enable us to gain insights into the experiences of both older people who experienced barriers or had the motivation to staying active and those who did not; we assumed that the most active and inactive might have very particular experiences such as severe ill health or unusual sporty pastimes. As these objective measures would form the basis of many of the epidemiological analyses of the overall project, we aimed to see in what way these might map onto people's everyday experiences.
A year on, we contacted all 27 participants, of whom 13 were able to attend a participant workshop. We also recruited 6 expert stakeholders through direct invitation from our existing networks and snowballing through personal recommendation, all of whom agreed both to attend the workshop and to be interviewed afterwards. These stakeholders were carefully chosen to represent a variety of local and national, practice and policy interests and comprised 2 county council public health officers, 3 regional civil society representatives and 1 national government representative either holding remits in physical activity and/or older people. We aimed for a small enough group of expert stakeholders not to intimate our lay participants.

| Data analysis
The initial verbatim interview transcripts were coded for thematic analysis, which involved familiarization with the textual data through repeated reading, identifying codes and synthesizing larger thematic categories. The coding was guided pragmatically by the research objectives but also allowed for inductive analysis of unanticipated topics or meanings. The ethnographic field notes served to triangulate the findings; they were not openly coded like the transcripts, but used to help at the later stage of analysis when codes were synthesized into categories, further exploring how categories played out differently in different situations or participants. We then further analysed our main findings (summarized in the concept of "activeness") 16  The study's 2 parts of qualitative data collection and the follow-up workshop and its embeddedness in a larger EPIC study enabled us to triangulate our findings in various ways. We were able to compare in what way the emergent qualitative typology matched the participants' objectively measured physical activity on which recruitment was based in particular the (see Table 1). We were also able to compare the initial interview responses with participants' comments on their synthesis and interpretation during the workshop.
The study received ethical approval from the NRES Committee South Central -Oxford C (14/SC/1047), and all participants gave their written informed consent. An amendment with separate written consent was sought for the participant workshop and follow-up interviews with expert stakeholders. Study results shared during the workshop were carefully anonymized.

| Developing types of activeness to understand differential barriers
We developed 3 motivational types of activeness, which included the disposition to be "exercisers," "out-and-about-ers" and the "sedentary/solitary" (see Table 1). We understand these types in the tradition of Weber's approach of ideal types, deriving them from, but not representing, social reality; instead, ideal types enable interpretation as analytical constructs with distinct explanatory characteristics and enable comparison within and across types. 18 This is important to note, as it means that participants could fall into several or all of the activeness types.
The "exerciser": Some of our participants described themselves as sporty and enjoying exercise, often in reference to their biographies. The quantitative measures for these participants show that all of our "exercisers" fell in the "active" category (objectively defined using accelerometer step counts; see Table 1). We should also note that most of our "exercisers" also fell into the category of "out-and-about-ers".
The "out-and-about-er": Many participants stayed active in a much broader way than by being explicitly physically active. This was particularly described as the motivation to be engaged in many social or intellectual activities. An "out-and-about" participant would, for example, list a variety of activities across a typical week, ranging from housework and gardening and meeting up with friends to looking after grandchildren or elderly parents, rarely missing out a day of the week in this tight scheduling. Some of these participants had less varied interests but were still leading busy work lives or were walking daily as their main means of transport. The participants quoted here used to do manual jobs in their working lives, and their perspective resonates with previous research that former manual workers may be more inclined towards purposeful activities in later life in preference to more leisure-based activities. 19 This has been described elsewhere as the "busy ethic", the strong motivation of continuing a work ethic into retirement. 20 However, in our sample, we could not find any difference between occupational classes in this respect, and those previously working in professional occupations also talked about the importance of staying busy.
The "out-and-about-ers" also experienced physical limitations as a main barrier to continuing lifelong activities. For example, a participant struggled to keep up bird watching in terrain that was now too uneven for them. Another explained that she stopped walking with a walking group because she would have had to keep up with the length of the walk and pace of the group. In the case of the "out-and-about-ers", though, their narratives were less about losing the ability to be physically active and more about losing out on social engagements and pastimes beyond sport. Interestingly, some of those classified by us as belonging to the sedentary/solitary type remained relatively physically active because their daily activities involved continued social responsibilities such as caring for grandchildren, siblings, parents or neighbours (see Table 1).

| Intervening in meaningful ways to promote active living
As described above, our typology reflects people's preferred ways of being active and does not necessarily map onto current activity levels. Common quantitative physical activity types or profiles tend to classify physical activity using (self-reported or objective) measures of the frequency or intensity of current behaviours; such types include "active," "inactive" and "sedentary" as well as more differentiated alternatives, for example "active but sedentary" (gym goers in sedentary jobs). 13 The distinction between present levels, past disposition and future aspirations seems crucial, however, as the barrier for such physical activity is not a lack of motivation but a lack of opportunity or capability. 21 In fact, the distinction in our typology between physical, social and intellectual motivations and activities can aid reflection on typical (formal) intervention strategies to promote active living.
"Classic" behavioural interventions tend to be framed within models that assume reasoned and planned behaviour based on the combination of a positive attitude, perceived social acceptability and capability to engage in a certain behaviour. 22 This maps somewhat onto our notion of motivation, a lifelong disposition, sense of self and a deliberate aspiration of how to spend the everyday 23 ; however, critical appraisals of behaviour change models suggest that attitudes and intention are not on a linear path towards individual behaviours, and barriers to certain behaviours often lie beyond personal attitudes in the social or physical realms. 14 More sophisticated intervention models acknowledge the importance of addressing the combination of capability, motivation and opportunity as these shape behaviour. 24 "Exercisers" in this study certainly enjoyed exercise all their life and did not need to be convinced thereof (eg, in social marketing campaigns); yet they experienced other barriers which prevented them from being active and would need to be addressed (eg, affordable or low-impact exercise opportunities, or age-specific classes). At first glance, the aim of typical exercise-based interventions is to make exercise more acceptable for "non-exercisers", for example by "disguising" workouts as dance (Zumba) or making exercise groups ageor gender-specific and therefore perhaps less intimidating. 25 Through the lens of our ideal types of activeness, however, they might not actually attract "non-exercisers" but simply target "low-hanging fruit" by providing more variety or adapted versions of activities for exercisers. To put simply, someone who considers herself "not a gym person" but prefers activities in nature will not necessarily feel more positive about a women-only gym. Targeting the population group of "women", here, fails to account for underlying varying dispositions.
While many "out-and-about-ers" might not be attracted by the offer of structured exercise, they might be more likely to appreciate safer walking routes or more convenient public transport to support their way of being active. Indeed, population strategies to promote active living now often include infrastructural improvements, for example providing pedestrian crossings, cycle routes or green spaces. 26 "Out-and-about-ers" with a busy schedule of activities might be supported by including physical activity in even sedentary engagements such as a theatre visit, if this could be reached by a safe and convenient walk. Such environmental interventions are sometimes based on the assumption: "build it and they'll use it" 27 ; if only physical or social environments were more supportive (eg, more pedestrian-friendly), people would realize that it is both feasible and, trying the new activity, even enjoyable. However, similar limitations might apply. Supportive environments for travel may speak best to those who travel. And intervention that promotes physical activity on a population level, for example through cycling networks, might create disabling environments for older people.

| Co-developing new strategies for staying active in later life
During our workshop, we invited participants to join our thinking. Moreover, the expert stakeholders were struck by the relatively low priority of "health" as an explicit motivator and summarized the workshop suggestions as "health by stealth", being engaged in physical activity for reasons other than trying to be physically active. "Out-and-about-ers" engaged purely incidentally in physical activity through taking part in their various interests and social relationships. However, the motivation to be active beyond physical activity as its main health benefit was even applicable to the exercisers; they seemed to gain enjoyment from movement, competitiveness or a challenge. These discussions of incidental physical activity then sparked ideas in some of them of "sneaking" active living programmes past funders "disguised" as dementia or social isolation programmes. There is an increasing recognition that public health strategies can focus on "nonhealth" outcomes such as sustainable travel and still be able to reap its health cobenefits. 28

| CON CLUS ION
Common quantitative physical activity research either classifies people into achieving particular physical activity levels or into profiles describing ability, socio-demographic characteristics or physical activity domains such as particular sports, transport and work or domestic activities. [10][11][12] To better understand how older people can be supported in staying active, we have developed a typology to help distinguish older people's motivations: those who have always enjoyed sport and exercise, those who are active in terms of social engagements and varied interests, and those who prefer more sedentary or solitary pursuits. We suggest that the limited success of many strategies in shifting population physical activity patterns might partly reflect a poor match with people's dispositions and aspirations; these interventions merely address the problem-low physical activity levels-without addressing specific barriers people experience in following their motivations due to challenges such as ill health or loss of activity companions.
To develop this motivational typology, we combined in-depth interviews with participant observation, for more in-depth understandings of participants' experiences, and workshop discussions for a combination of informal and formal reflections on our findings. Important for us was the participatory nature of this study, to allow participants to show and share with us their common activities and pastimes, and the social and physical environments in which these take place. The participatory workshop then enabled us to share our preliminary findings and emerging conceptual ideas, gain feedback for their usefulness from stakeholders in policy and practice and encourage dialogue between lay and expert participants.
Our participants identified ideas that can speak to a variety of dispositions and respective barriers, for example by enabling both physical and intellectual activity, solitary or social pursuits, and often promoting physical activity as a secondary "co-benefit". Intervention strategies increasingly acknowledge these issues, for example by adding "opportunity" to behaviour change models 24 and increasingly targeting broader concepts of physical activity to reach out beyond exercisers. 26 Including participants more effectively in developing policy and practice, and recognizing the importance of people's biographies and social worlds, might be equally important in identifying meaningful active living strategies.

ACK N OWLED G EM ENTS
We thank all participants, as well as Guy Shefer, who was the second ethnographer for the phase 1 data collection and analysis.

CO N FLI C T O F I NTE R E S T S
The authors declare no conflict of interests.

AUTH O R S' CO NTR I B UTI O N S
CG, SG and DO were involved in the study conception and design. SG and DO obtained the funding. CG collected the data. All authors were involved in the analysis and interpretation of the data. CG drafted the article; JP, SG and DO revised it critically for important intellectual content; all authors approved the final version.