Multiple influences on participating in physical activity in older age: Developing a social ecological approach

Abstract Background Evidence of the benefits of engaging in physical activity (PA) is strong, yet the number of older adults meeting the recommended 150 min/wk is low. Policy to increase uptake and adherence has focussed on the health benefits of PA, but may not be the most successful approach. Objective This qualitative study sought to ask older adults what the components of a successful intervention to promote PA would be, by asking active older adults what motivated them to be active and asking inactive older adults what might encourage them to change. Design, setting and participants Focus groups and semi‐structured interviews were held with 60 community‐dwelling older adults, aged 50‐87 years. Framework analysis was used to analyse the data, and themes congruent with a social ecological model of behaviour change were developed. Findings Five themes emerged that influenced PA engagement at multiple levels: individual; interpersonal; perceived environment; community or organizational; and policy. PA engagement was determined by attitude or health status for some participants, but for the majority, PA being enjoyable, sociable, affordable, accessible, flexible and seasonal were more important than the health benefits. Discussion and conclusions A social ecological model is presented, highlighting the fact that both motivated and unmotivated older adults need to have a range of appropriately labelled, appealing and accessible activities to choose from when thinking about engaging in PA. Policymakers and practitioners need to ensure that their offers of activity sessions are easy to access and easy to remain involved in.


| INTRODUCTION
The health benefits of physical activity (PA) are well documented with higher levels and greater frequency of PA being associated with reduced risk and improved health in a number of key areas. [1][2][3][4] Improvements in mental health, well-being and cognitive function are also associated with regular PA. [5][6][7][8] Despite these health benefits, PA levels amongst older adults remain below the recommended 150 min/ wk. 1,9 International and national health policies have focused on improving health by providing PA guidance to the older population [10][11][12] and by highlighting to those working with older adults the need to incorporate increased amounts of PA into everyday life. 1,13 Contemporary policy developments in the UK have sought to involve people in improving their own health and well-being, to maximize the opportunities for better health outcomes. 14,15 Participative approaches have been used in the management of long-term conditions, nutrition and addiction, 16,17 and there are suggestions that this could be fruitful in other areas. A literature review examining the involvement of older adults in the design, delivery, implementation and promotion of interventions to increase PA found only ten studies in which older adults played an active and prominent role. 18 Of these, only four studies demonstrated that older adults were involved in designing and promoting an intervention, based on factors that would appeal to their age group. [19][20][21][22] In the wider literature, there are a number of examples of older adults developing individual action plans and coping plans within interventions, 23 but not in designing the structure of the overarching interventions, which are delivered within particular environmental, organizational and policy structures. 24 Social-ecological approaches to public health interventions provide a framework for understanding the importance of the dynamic interrelations between a person and their environment, and the context within which they exist, recognizing the complexity of human situations. 25 Whilst recognizing the importance of psychosocial factors on behavioural change, such as knowledge, attitudes, beliefs, selfefficacy and social support, [26][27][28][29][30] Social Ecological Models (SEMs) also include consideration of influences at organizational, environmental and policy levels. 24,25,31 The aims of this qualitative study were to (i) explore the views and experiences of older adults in relation to successful PA interventions and (ii) to develop recommendations with older adults for a population level PA intervention to promote uptake and adherence. The research question was as follows: "What do older adults think the essential characteristics and preconditions are for the delivery of a successful intervention to promote physical activity amongst their age group?" The study follows a pragmatic approach to consider influences on PA engagement at multiple levels, [32][33][34][35] leading to the development of a SEM for promoting PA amongst older adults.

| Participants
Community-dwelling older adults, aged 50 years and older, were recruited through non-government organizations and community groups in West Yorkshire, UK, being purposively and snowball sampled to recruit participants with different experiences of participation or non-participation in PA. Due to limited resources, people who did not speak English, or who had other significant communication difficulties requiring a translator, were initially excluded. An amendment was approved to include two participants who spoke Urdu. Recruitment continued until no novel data were produced. 36

| Data collection
Prior to the recruitment of participants, a public involvement consultation exercise was undertaken with 12 older adults to formulate and clarify the questions that would be asked in focus groups and interviews. 38 Topics included participants' understanding of PA; the PA undertaken by participants; their reasons for uptake, adherence and termination of PA; their views on group and individual PA, including environmental and social factors; and their views on the promotion of PA amongst their age group. The majority of participants recruited opted to take part in focus groups between October 2011 and April 2012. Over the same period, semi-structured interviews were conducted with those who preferred to meet 1:1 or were unable to attend focus groups. Data collection took place in community buildings and participants' homes.

| Data analysis
Given the relevance of this study to the national policy framework regarding promotion of PA, prevention and healthier communities, the framework approach to data analysis was used. 39 This approach facilitates systematic qualitative analysis, summarizing and classifying data within a thematic framework. 40 The focus group and interview recordings were transcribed verbatim and initially coded using NVivo9 41 by the first author, before meeting with the other authors.
Classification began following the first three data collection events and continued throughout the data collection period. Through discussion, all three authors developed a thematic framework containing key themes and subthemes. This framework was systematically applied to all of the data, being modified and refined through five phases of review and analysis. All authors met to discuss and develop the framework at regular intervals, classifying and synthesizing data, reviewing

| RESULTS
Eleven focus groups and 12 interviews were conducted. Participants were aged between 50 and 87 years, with the majority being female.
The gender, age, ethnicity and level of PA of the 60 participants recruited are reported in Table 1. Participants were described as physically active if they met the current recommendations. 3 Five themes emerged that influenced PA engagement at multiple levels of the SEM: individual/intrapersonal; relationships/interpersonal; perceived environment; community/organizational; and policy levels. The SEM developed from the findings is presented in Figure 1.
Exemplar quotes are included in the main text, with additional supporting quotes presented in Table 2. Quotes from participants meeting the recommendations for PA are indicated by the word "Active" preceding their gender and age.

| Individual/intrapersonal level
Here, participants talked about their own ability, motivation and belief in the health benefits of PA. The influence of participants' personalities and lifestyles was described and reflected upon.

| It's more about me
There was a range of engagement in PA amongst participants. From very little activity (walking to the shops, travelling to a social activity) to PA dominating their week (active in groups, or on their own, on every day of the week). The most commonly reported driver for PA within the very active subgroup was that they had always been active;

| My own ability
Feeling fortunate to be fit was expressed in three of the focus groups.
"The way I look at it, I'm lucky to be able to get about the best way I can. So I'm just glad about that." (Female, 78) In contrast, many participants spoke of their frustration at ill health and poor mobility restricting them. For some, this frustration was increased because they had previously been active (Quote 3).

| It's good for me
Whilst the health benefits of PA were not spoken about by many participants, for some they were a major driver.

| Relationships/intrapersonal level
Here, participants described the social influences and benefits that they experienced.

| Sociable
The social element of PA was rated highly by many. Meeting new people, forming friendships and socializing with like-minded people were often more attractive than the activity itself (Quote 6). Companionship and company were particularly important factors for participants who lived alone (Quote 7). Several talked about forcing themselves to go out to combat loneliness. "I live on me own and I get fed up on me own.

| Perceived environment
Here, participants talked about the need for sessions to be easy to access, not just in terms of physical access (on a bus route, with good car parking, no steps), but also through atmosphere and composition.

| Affordable
The majority of participants were retired with pensions of varying amounts. Only five participants were in employment and two were on work-related benefits. Difficult decisions about how to spend limited income were reported. "I'd go more often, but it's a bit, money's a bit tight. I find at the moment that I would like to join a lot more things, but they all cost." (Active Female, 76) The loss of subsidised and free activity sessions, such as adult education classes and swimming, meant that some participants' opportunities for activities had been cut off. Paying for a set of classes in advance was out of the question for most, who preferred a "pay-as-you-go" approach, which enabled them to budget and adapt (Quote 10).

| Accessible
Being able to engage in PA very close to home was important. Access to venues needed to be as easy as possible. "Easy access. home. There did appear to be a relationship between the outdoor environment and the amount of PA undertaken. Of the 15 participants who spoke of the attractiveness of the local area, ten were regular walkers and five were motivated to walk, although not regularly.

| Weather, seasons and time
Many reported being less active in the winter, less willing to go out after dark and in bad weather (Quote 13

| Community/organizational level
Here, participants reported various challenges for those encouraging older adults to engage in PA.

| Appealing to men
According to participants attending group PA sessions, women domi-

| Appealing to all ages
There was some frustration amongst participants about the labelling of activities for "over 50s." There was a perception that the age range was too great. Some younger older adults did not identify themselves in the same age group as those in their 70s. The "over 50s" term was described as exclusive, preventing engagement from many who did not identify themselves by their age. "So once you start putting ages, you're kind of ruling some people out who'd probably try it, you know." (Active Female, 64) Many people over 50 are still in work, and the "over 50s" classes were often run during the day. One participant was most emphatic that they were "retired people's classes," rather than for those over 50.

| Appealing to the uninterested
In the discussions about how to motivate the uninterested to engage, there were more questions and recounts of failed attempts than answers and suggestions. Using a variety of methods and resources to promote activities was suggested, with word of mouth being thought to be the most effective tool. "You need it more than one way as well, don't you? A personal invitation and perhaps something to give them to remind them." (Active Female, 64) Thinking about the target audience, and using language that people can identify with, was particularly important for the South Asian Ladies' group "Gup Shup" (Chit Chat), as was conducting the activity sessions within their own locality.

| DISCUSSION
This article addresses the applied public health issue of how to encourage greater levels of PA amongst older adults. To help us achieve practical goals, we have framed this article within a social ecological approach. 24 We recognize that we could approach this issue from a variety of theoretical perspectives and that each may highlight different factors. However, SEMs have been used successfully to promote PA because individuals, with their own unique motivations, engage in activity that takes place in particular settings and contexts. 30 SEMs permit inclusion of a broad range of variables, from different domains, and recognize the importance of the dynamic, multidirectional influences between people and their physical, social and political environments. 25 The interaction of personal, social, environmental, organizational and political variables on health is of particular relevance to policymaking and public health implementation.
The multilevel influences on engagement in PA uncovered by this research are presented in Figure 1  Participants' experiences of these factors influence how they feel and act and thus influence the relationships that they have at the interpersonal level. There is a dynamic, multidirectional relationship between the SEM levels. 25 The importance of affordable activities has a strong evidence base, to which this study contributes. 49,50,53,54 Within this study, this has been linked to safe and easy access to activities through public transport, facilitated by the free bus pass for those of State Pension age. 51 This study confirms the importance of ensuring that PA is enjoyable, as very few people stated that they were active for the health benefits alone. 50,[52][53][54] The outdoor environment has been shown to encourage people to be more active in their local areas, 29,55 but this is mitigated by the influence of poor weather and dark nights on engagement in PA and attendance at activity sessions. 43,46,47,51,56,57 This study has found that engagement in PA is seasonal, with activity levels increasing in the warmer months. Some PA interventions are season-specific, such as outdoor bowling, and alternatives may need to be offered to maintain PA levels throughout the year. There should be variety within activity sessions, to ensure that participants remain interested and engaged. 44 These are considerations for those organizing PA, leading us into the next level.
At the community, or organizational level of the SEM are findings regarding image and presentation. Activity sessions should be labelled appropriately, in order for people to identify with the activity on offer.
Using the "over 50s" label may not be the best way to promote PA amongst this age group, as few people identified as "over 50." 58 The age range from 50 years and upwards is too large to be meaningful to every person within that range. Use of the label "over 50" has not been widely reported as a barrier to engagement in community-based activities for older adults and this warrants further consideration. The action research study following on from this qualitative study considered this issue in more depth, as the researchers sought to find a way of appealing to local community members over the age of 50.
Consideration should also be given as to how to appeal to different subgroups of the older adult population, to men and the "un-  61 Other studies have shown that Muslim women are deterred from attending mixed-gender activities due to their own or family concerns over modesty and gendered norms. 62 Therefore, to ensure participation of BME groups, facilities congruent with their practices and beliefs need to be developed. 30,62 A systematic review of interventions to promote PA amongst people with severe mental illness showed that lack of support was a major socio-economic barrier to engagement. 63 Organizers and promoters of PA interventions need, then, to ensure that appropriate support is available to facilitate engagement. The challenge of appealing to those regarded as "uninterested" should also be addressed when considering image and presentation. This study confirms the difficulties of attracting older adults who most need to increase their PA levels 64  Concerns about the sustainability of activity sessions were raised by participants, who wanted to see local government departments provide both financial and practical resources to support activity sessions.

| Strengths and limitations
The participants in this study were recruited through existing community groups and at an "activity showcase" event. Snowball sampling led to recruitment of a further 23 participants. Each participant was either physically active, or involved in some kind of social group activity, or both. Reference was made to family members or friends who did not engage in any activity (social or physical), but they were not directly involved in this study. The study was geographically bound in the north of England, but the methods described mean that it could be replicated elsewhere.
The lead author had worked and volunteered in the area where the study was conducted for more than 15 years at the time the study was conducted. Participants could recognize a long-standing commitment to older adults in the locality, with knowledge of the local situation.

| Recommendations for policy and practice
Policymakers can do much to improve the promotion of PA at multiple levels of the SEM. As few participants were driven to take part in PA to gain physical health benefits, more focus should be given to the social benefits and the fun, laughter and enjoyment that can be obtained. The labels applied to PA promotion should be as open and inclusive as possible. Ease of access should also be considered, in terms of cost, location, a friendly welcome and timings that reflect seasonal changes. In particular, the temptation to remove the free bus pass for people of State Pension age should be resisted. Local government departments and health services should consider the cost benefits of providing resources to support community-based activity sessions, not least in creating enjoyable environments, because the health benefits of PA are well-known.

| Recommendations for further research
Findings from this study suggest that the "over 50s" label is perceived as exclusive, yet it is common for initiatives to use this label. Further research into older adults' views on the labelling of groups and activities is required, to see whether this is a major barrier to engagement.
Further research is also needed into how to appeal to those perceived as marginalized, disengaged and uninterested in PA. Participatory approaches would be most useful here, to understand the difficulties and develop potential solutions from the perspective of those affected.

| CONCLUSION
This study found that, in some cases, there are personal attributes that override the extrinsic factors that policymakers and local organizations can have some influence on. Poor health, and mobility, feeling the need to be active (or not), and being motivated to engage in activity can be the primary driver for PA engagement. Not everyone will respond positively to targeted advertisement of PA. However, through ensuring that activities are flexible and adaptable, making sure that people feel welcome regardless of ability, could help to encourage older adults to try activities that they might otherwise rule out. Ensuring that activities on offer have maximum appeal is a key element of increasing engagement. Interventions that focus on individual-level predictors of engagement require attractive activities to signpost older adults to. The dynamic relationship between influences at all five levels of a SEM will assist in the development of interventions that take into account optimum conditions for engagement.
The focus of public health interventions to promote PA has largely stressed the health benefits. 1,9 However, adopting a multilevel approach to promoting engagement in PA, by focussing on the social elements of engaging in activity, together with ensuring that it is easy to take part, both on a practical and emotional level, may do more to encourage older adults to meet the recommendations of 150 minutes of moderate intensity PA over the course of a week. 1,9