Predictors of intentions of adults over 35 years to participate in walking sport programs: A social‐ecological mixed‐methods approach

There is a growing need to identify acceptable and feasible opportunities to engage adults over 35 years in physical activity. Walking sports may be a potential means to engage adults in sport; however, there is limited evidence regarding appeal and feasibility to support its implementation and delivery. Using a two‐step mixed‐methods approach, we aimed (1) to quantitively identify significant predictors of intentions of adults over 35 years to participate in walking sports and (2) to understand why and how these identified predictors may be contextually relevant to the target group. In phase one, 282 adults over 35 years (Mage = 46.08, SD = 9.75) without prior experience of walking sports completed an online questionnaire assessing personal, psychosocial, program‐related, and environmental predictors, and intentions to participate in walking sports. Hierarchical multiple linear regressions showed that perceived health status, attitudes, subjective norms, and distance of venue were significant predictors of intentions. In phase two, interviews with a subset of 17 participants indicated that, when implementing walking sport programs, program labeling, fear of the unknown, and individual differences in the appeal of walking sport warrant consideration. Together, these findings offer insight into the complex interplay of personal, psychosocial, program‐related, and environmental predictors of adults' intentions to participate in walking sports. Addressing these elements of a walking sport program would make such programs more appealing to potential participants, and ultimately, more feasible and sustainable to conduct in the long run.


| INTRODUCTION
Global trends suggest that physical activity levels decline with age. 1 One-third of the adult population is considered insufficiently active (i.e., performing less than 150 min/ week of moderate physical activity). 1 For instance, in the Americas, Eastern Mediterranean, and European regions, over half the adult population over 60 years of age is insufficiently active. 1 Physical activity has the potential to confer benefits across physical, psychological, and social facets of health, 2 and insufficient physical activity has been noted as a key risk factor for health issues in aging adults. 1 Therefore, it is imperative to identify appropriate physical activity options for aging adults, and encourage them to be more physically active, as a primary prevention tool for chronic illness.
Sport is a type of physical activity that involves physical or motor skills adhering to a specified set of rules while competing against other individuals or teams. 3 Meta-analytic evidence highlights that, for adults, sport participation is linked with physiological benefits for cardiovascular, metabolic, and musculoskeletal health. 4 A recent meta-analysis has also indicated that, for adults, sport participation is associated with psychosocial benefits, such as improved subjective well-being and social connectedness. 5 Compared with other forms of sport (e.g., individual sports), team sport participation in adulthood may confer unique benefits-the social interaction and social networks afforded by team membership are associated with social benefits such as social integration and belongingness. 6 For instance, participation in "Back to Netball," a UK-based community netball program for adults, fostered a sense of relatedness among participants and competence for netball participation, which could, in turn, enable long long-term engagement in physical activity. 7 However, sport participation markedly declines with age, 8 perhaps because of the lack of appropriate sport opportunities for aging adults, highlighting the need to cater to this age group. 9

| Walking sport programs
Recently, adapted sports have gained traction. These sports involve altering specific aspects of traditional sports (e.g., football/soccer, netball) to provide suitable opportunities for individuals who would need sports to be adapted to participate. Walking sports are a specific type of adapted sport where, instead of running (as is typical in a traditional sport), participants are restricted to walking-the focus, therefore, is on positioning, motor skills, and team work, rather than cardiovascular capacity. 10 Typically, when compared with traditional versions of the same sports, walking sports are adapted to be noncontact, thereby lowering the impact and risk of injury. 11 Therefore, walking sports may be particularly appropriate for, and appealing to, an inactive or older section of the population, who may wish to benefit from increasing their physical activity, but cannot keep up with the high intensity of traditional sports. 12 Although walking sport programs have gained popularity in recent years, there is limited empirical evidence to support the implementation and delivery of appealing, feasible, and sustainable walking sport programs. 12 Preliminary investigations in the UK suggest that walking football (soccer) and walking netball programs are feasible, engaging, and sustainable when delivered to adults within community settings. 10,11,13 The feasibility and acceptability of walking netball programs were attributed to the participatory approach to implementation, which permitted adaptations to be made to cater to the needs of the target population. 13 The social connections resulting from team membership were also considered particularly important in sustained, long-term engagement. 11 However, the informativeness of these investigations 10,11 are limited by the small sample sizes with a small intervention duration.
Recent research has explored the experiences of participants and sport organizations involved with walking sports. 10,14,15 Regarding initial attraction to walking soccer, participants retrospectively recalled that their sporting history, positive perceptions of walking soccer, and high self-efficacy for walking soccer were key enablers, whereas low perceived physical capability and social exclusion were potential barriers to participation. 10,14 Furthermore, positive experiences within these programs, social connectedness, and perceived health benefits were highly influential in sustaining participation. 10,15 Given the multiple potential health benefits of walking sport participation, and the declining trends of physical activity and sport participation in aging adults, 8 there is a need to ascertain the key predictors of walking sport participation to identify means to encourage participation in such community-based sport programs.

| The social-ecological approach
The social-ecological approach suggests that adults' physical activity choices are influenced by several predictors at personal, psychosocial, environmental, organizational, and policy levels. 16 These predictors extend from ones that are most proximal to the individual (i.e., personal) to ones that are distal contextual predictors (i.e., policy). The social-ecological model is relevant to understanding adults' walking sport participation. 13 In this section, we provide a brief overview of some key predictors and theories relevant to physical activity and sport participation in aging adults at each level of the social-ecological model.
At the personal level, considered to be the most proximal to an individual, a wide range of demographic characteristics (e.g., age, gender, ethnicity, relationship status, health status, prior sport experience) have generally been associated with physical activity and sport participation. 9,14,17,18 At the psychosocial level, the theory of planned behavior has been frequently used to explain individuals' intentions toward specific behaviors (including physical activity). 19 This theory focuses on three key variables-attitudes (i.e., the extent to which an individual has a favorable or unfavorable evaluation of the behavior), subjective norms (i.e., the belief regarding others' approval of, or participation in, the behavior), and perceived behavioral control (i.e., the individual's perception of ease/difficulty of performing the behavior). 19 Together, these variables are shown to predict 44% of the variance in behavioral intentions. 20 Social support has also been identified as instrumental in engaging adults in physical activity and sport. 21 At the program level, extant literature has indicated a wide range of program-specific predictors that may influence adults' engagement in walking sport programs, as well as physical activity more broadly. Specifically, evidence suggests that adults may have varied preferences for organized (rather than unorganized) and competitive (as opposed to recreational) sports. 8 The day(s) of the week, cost of participation, and proximity of the venue have also been associated with sport participation. 9,14 Furthermore, the age range and genders of other participants, and group size were also implicated in adults' interest in a specific sport program. 13,18 At the environmental level, aspects of an individual's local neighborhood have been found to influence sport participation, with low perceived safety and high traffic congestion inhibiting sport participation. 9,16 However, the relative importance of each of these predictors across the social-ecological levels when considered concurrently is unknown, and warrants further attention.

| Gaps in the evidence
Extant research is limited in scope for at least two key reasons. First, prior research 14,15 has solely considered the perspectives of, and thereby catered to the needs of, existing walking sport participants. These investigations, utilizing retrospective narratives to understand walking sport initiation, carry potential bias in participants' recollection of information, and may be limited to the views of individuals who ultimately engaged with walking sports. To expand the delivery of walking sport programs, there is a need to attract new participants (with diverse demographic features) to such programs, and to consider their perspectives, and cater to their needs. 18 Second, the foci of past investigations have been limited to personal and psychosocial predictors pertinent to walking sport participation. There is a need to consider the extent to which program-related and environmental predictors may also be relevant to walking sport participation. A socialecological approach would allow for the inclusion of program-related and environmental predictors with personal and psychosocial predictors, enabling a multilevel understanding of the predictors of aging adults' intentions to participate in walking sports. Such an enquiry could lend insights that may be particularly beneficial for the growth and wider appeal of walking sport programs.

| The present study
To address the two aforementioned limitations (i.e., the focus on existing participants and on only some of the levels of the social-ecological model), our mixed-methods investigation focused on the perspectives of individuals who fit the target population in terms of age, but have no prior experience with walking sports (i.e., potential participants). We sought to build an understanding of the facets of walking sports that are appealing (or less so) to potential participants, and develop recommendations to support the future implementation of walking sports that would be feasible and sustainable for sport organizations to deliver in the long run. Given the hypothetical nature of the behavior, and our focus on the appeal of walking sport programs, we considered intentions rather than actual behavior as our primary outcome. First, as previous research suggested that each level of the social-ecological model may be associated with physical activity, and specifically walking sport participation, we aimed to quantitively identify significant social-ecological (i.e., personal, psychosocial, program-related, and environmental) predictors of intentions to participate in walking sports for adults over 35 years. We expected that each level of the socialecological model would uniquely contribute to explaining intentions to participate in walking sports. Second, we aimed to understand why these identified predictors may be contextually relevant to the target group. Qualitative methods enabled us to situate the quantitative responses within respondents' personal circumstances and *We use the term 'predictors' to mean statistical prediction, as originally used by Cohen in his seminal work on regression analysis (see Cohen et al. 43 ). We do not intend to imply causality, given that quantitative data was obtained using cross-sectional self-report questionnaires.
experiences that may not have been captured through quantitative measures.

| Study design
This mixed-methods study was part of a larger research project that investigated the acceptability, feasibility, and sustainability of community-based walking sport programs. Our overarching aim was to better understand real-world challenges in relation to physical activity and health promotion through walking sport participation in the context of an aging population. The research team encompassed individuals of varied expertise (i.e., qualitative, quantitative, mixed methods, and implementation research). Our philosophical positioning is dialectical pluralism-a meta-paradigmatic stance that considers multiple paradigms such that these paradigms can be integrated to provide a broad understanding via multiple perspectives and values. 22 Ontologically, dialectical pluralism is pluralistic, embracing multiple realities; epistemologically, this positioning refrains from relying on a single standpoint concerning knowledge production. 22 Instead, it is imperative for researchers to accept multiple ontologies, which may all offer noteworthy insights, and ascertain what is epistemically pertinent to a particular research project. 22 In our present research, we selected paradigms to align with each of our research objectives. We had two key objectives: (1) to quantitively identify significant predictors of intentions to participate in walking sports and (2) to understand why and how these identified predictors may be contextually relevant to the target group. Hence, our present research is framed by the realist paradigm and a positivist approach (to address the first objective), and a relativist paradigm and an interpretivist approach (to address the second objective).
In line with our philosophical stance, we adopted an explanatory sequential mixed-method approach, comprising of a quantitative survey to address the first objective (phase one) followed by qualitative interviews with a subset of respondents to address the second objective (phase two). This sequential approach was chosen for two reasons: (1) to inform interviewee selection from the survey and (2) so that the significant predictors identified from the survey could be further explored in the interviews. Correspondingly, the weighting of these two phases was unequal-there was greater emphasis on the qualitative data as the team agreed that it was important to understand the meaning behind the quantitative data, and hence, valued individual perspectives over generalizing quantitative findings. 23

| Participants
In phase one, 311 adults aged between 35 and 84 years (49.8% male, 50.2% female, M age = 46.08, SD = 9.75) completed an online questionnaire. Inclusion criteria required that all respondents were 35 years or older, based in Australia, and had never participated in walking sports at the time of data collection. We included adults aged 35 years and older, as individuals over 35 years of age may begin to participate in Masters' Sports, while a large proportion of adults disengage from sports. 24 Respondents varied in their level of physical activity; 27.3%, 33.1%, and 39.5% of participants reported a low, moderate, and high level of physical activity, respectively, with 27.3% of respondents insufficiently active. In addition, 67.5% of respondents reported having previously played traditional forms of sport. There were 119 (1.20%) missing values noted across 22 variables in the dataset, which, upon inspection, appeared to be missing at random. This has been noted as "inconsequential" to the statistical analysis. 25 Therefore, we performed the regression analyses on the subset of complete cases (N = 282). Power analysis using G*Power indicated that our sample size is sufficiently powered (90% power, α = 0.01) to detect effects as small as 0.15 for incremental step variance in multiple linear regressions. Respondents' demographic information is presented in Tables 1 and 2.
Participants for phase two were identified using criterion-based purposive sampling to select individuals of wide-ranging demographic features and perceptions of walking sports to ensure that we considered a variety of experiences. Eligible individuals were those who scored within the top and bottom 10 percentiles on the measure of intentions in phase one. In all, 17 individuals (8 males, 9 females, M age = 45.76, SD = 8.15) across six states in Australia (i.e., Queensland, South Australia, Western Australia, Tasmania, New South Wales, and Victoria) completed phase two. In all, 11 of these individuals reported having participated in traditional sports (either in the past or at present). The wide range of participants interviewed allowed us to consider diverse perspectives, contributing to the richness of data and information power. 26 Interviewees' demographic information is presented in Table 3.

| Procedure
Once institutional ethical approval was obtained from an Australian university, participants were recruited through social media (i.e., Facebook, Twitter). Participants were given a link to provide informed consent and complete an online survey via Qualtrics, which took approximately 25 min to complete. The survey measured personal, psychosocial, program-related, and environmental constructs that were identified in previous research 13,16 as potentially relevant to intentions to participate in walking sports (and physical activity broadly). The full questionnaire is presented as Appendix S1. Following the analysis of quantitative data, eligible respondents were contacted to arrange a follow-up interview. We interviewed consenting participants (50% acceptance rate) via Microsoft Teams or by phone, in adherence with COVID-19 regulations at the time of interviewing. Semi-structured interviews explored participants' experiences with physical activity (e.g., "Can you tell me a little more about your experience with participating in physical activities?"), the appeal of walking sports (e.g., "What is your impression of walking sports?"), the significant predictors identified in phase one (e.g., "To what extent would your friends and family's opinions on walking sports influence whether you choose to participate?"), and barriers to walking sports (e.g., "What are some things that might make it difficult for you to participate in walking sports?"). The full interview guide is presented as Appendix S1. Interviews were recorded and manually transcribed verbatim. On average, the interviews lasted 36 minutes (ranging from 24 to 58 min).

| Measures
Measures were selected to assess relevant predictors of people's intentions to participate in walking sports at each level of the social-ecological model.

| Personal predictors
Participants were asked to report their age, gender, education level, employment status, income, relationship status, ethnicity, and primary mode of transport. The education and income categories were aligned with those of the national census. 27 One item along a 5-point Likert scale (1 = poor to 5 = excellent) constructed for the purpose of this study was used to measure respondents' perceived health status (i.e., "In general, would you say that your health is…"). A further question constructed for the purpose of this study probed participants' sport history (i.e., "In your adult life [since the age of 18], have you participated in any sport previously [for a minimum of 6 months]?"). As physically active individuals may be more inclined to initiate walking sport participation than their inactive counterparts, 14 the International Physical Activity Questionnaire-Short Form (IPAQ-SF) 28 was used as a measure of physical activity level. Responding to seven items, participants were prompted to recall how physically active they had been in the past 7 days (e.g., "During the last 7 days, on how many days did you walk for at least 10 minutes at a time?"). Together, these responses provided a measure of physical activity in terms of metabolic equivalents, and allowed for classifying participants' activity level as either "high," "moderate," or "low." The IPAQ-SF scores have been shown to be reliable and valid for use with a wide range of adult populations. 28

| Psychosocial predictors
Items measuring attitude toward walking sport, subjective norms, and perceived behavioral control were constructed in accordance with recommendations by Ajzen. 29 All items were preceded by the stem "If a walking sport program is made available to me in the near future…," reflecting that walking sport programs might not have been a currently available option for some of the respondents. Two dimensions of attitudes (i.e., affective and evaluative) were measured using eight questions employing a 7-point semantic differential scale ranging from −3 to 3. These items were stemmed by "participating in walking sport would be…." Adopting the approach utilized by Thøgersen-Ntoumani, 30 four items each measured affective attitudes (enjoyableunenjoyable, boring-interesting, pleasant-unpleasant, and stressful-relaxing) and evaluative attitudes (useful-useless, harmful-beneficial, wise-foolish, and bad-good). In the present study, scores from this attitude scale demonstrated high internal reliability (Cronbach's α = 0.91). Subjective norms and perceived behavioral control were measured with a 7-point Likert scale (1 = strongly disagree to 7 = strongly agree). Following Ajzen's 29 recommendations, we wrote four items each measuring subjective norms (e.g., "My family would want me to participate in walking sport") and perceived behavioral control (e.g., "I am confident that I could participate in walking sport if I really wanted to"). In the present study, scores from the measure of subjective norms and perceived behavioral control demonstrated high (Cronbach's α = 0.84) and marginal (Cronbach's α = 0.64) internal reliability, respectively.
The Physical Activity Social Support Scale 31 was used to measure social support. We excluded the information, validation, and companionship support subscales, as they were less relevant to aging adults and the nature of walking sports. We used the emotional support and instrumental support subscales which comprised four items each assessing the psychological and practical support that may be necessary for the target population to engage in physical activities. For each item, participants responded to a 7-point Likert scale (1 = never true to 7 = always true). We modified the stem of the items to allow participants to reflect on a hypothetical, rather than actual, situation, so that the items were relevant to everyone (i.e., "If needed, I would have someone who…"). Sample items for emotional and instrumental support included "could provide reassurance" and "could provide help traveling to participate in walking sports" respectively. In the present study, scores from this social support measure demonstrated high internal reliability (Cronbach's α = 0.92). This measure has been used with an adult population. 31

| Program-related predictors
We identified potential program-related predictors of intentions to participate from existing literature on physical activity and walking sport programs. 15,16,30 For each of the nine identified predictors, we asked participants to rate the extent to which each predictor may influence their decision to participate on a 7-point scale (−3 = unlikely to participate to +3 = very likely to participate). These items encompassed features of the sessions (i.e., coached sessions, turn up and play, social/recreational, competitions, weekdays, and weekends) and participants (i.e., age, gender, group size) that may be relevant to consider. On a sliding scale, participants were also asked to indicate the maximum fee they were willing to pay per session (ranging from AUD$0 to AUD$20) and the maximum distance they were willing to travel to participate in walking sports (ranging from 0 to 50 km).

| Environmental predictors
We used an adapted version of the Physical Activity Neighborhood Environment Survey (PANES) 32 to measure environmental predictors. We selected six items from T A B L E 3 Demographic information for interviewees.

Pseudonym
Age Gender PANES, measured on a 4-point Likert scale (1 = strongly disagree to 4 = strongly agree), that were directly relevant to aging adults and walking sport participation. Two items each measured aspects of crime (e.g., "The crime rate in my neighborhood makes it unsafe to go on walks during the day") and traffic (e.g., "There is so much traffic on the streets that it makes it difficult or unpleasant to walk in my neighborhood"). One item each was used to measure access to recreational facilities (i.e., "My neighborhood has several free or low-cost recreation facilities, such as parks, walking trails, bike paths, recreation centers, playgrounds, public swimming pools, etc.") and the observation of physically active people in the environment (i.e., "I see many people being physically active in my neighborhood doing things like walking, jogging, cycling, or playing sports and active games"). In the present study, scores from this scale demonstrated adequate internal reliability (Cronbach's α = 0.70).
We asked participants to report their residential postcode. We mapped respondents' postcodes onto Australian Bureau of Statistics data on Socio-Economic Indexes for Areas. Specifically, we used the Index of Relative Socioeconomic Advantage and Disadvantage (IRSAD)-using deciles, this metric summarized the social and economic condition of households within a specific area (1 = most deprived to 10 = least deprived area). 27

| Intentions
Three items measuring intentions were also constructed adhering to the recommendations of Ajzen. 29 These items (e.g., "I would plan to sign up") were preceded by the stem "If a walking sport program is made available to me in the near future…," Items were measured along a 7-point Likert scale (1 = strongly disagree to 7 = strongly agree). In the present study, scores from this measure demonstrated good internal reliability (Cronbach's α = 0.81).

| Data analysis
In phase one, we analyzed quantitative data using SPSS 28.0. We used hierarchical multiple linear regressions; given that prior investigations focused primarily on personal and psychosocial predictors of walking sport participation, we wished to consider whether program-related and environmental predictors would explain variance in intentions beyond personal and psychosocial predictors. Therefore, the most proximal (i.e., personal) predictors were entered at step one, followed by the more distal psychosocial predictors at step two. Furthermore, distal program-related predictors were entered at step three as these were specific to walking sport participation, and finally, general environmental predictors were entered at step four. Given the large number of predictors within our analyses, we adopted a cut-off of p ≤ 0.01 as indicative of significance.
In phase two, we analyzed qualitative data iteratively and deductively using reflexive thematic analysis. 33 To promote data familiarization, HS transcribed all interviews verbatim and reviewed transcripts, making reflection notes throughout the process. HS assigned initial semantic codes-this was done deductively as the aim was to provide further insight into why certain predictors (identified in phase one) were relevant to respondents' intentions to participate in walking sports. 33 Codes were revisited to interpret meaning and patterns in the data, facilitating the development of initial themes. The process of theme development was conducted iteratively, identifying patterns in the data while revisiting the statistical findings to ensure the themes developed explained these statistical findings. This enabled us to cluster codes around some key central organizing concepts. Through the theme development process, EQ served as a "critical friend," improving the richness of the data. For instance, the theme "Promoting achievability of the activity" was originally called "Importance of self-efficacy for the activity." However, upon discussion, we agreed that the findings extended beyond self-efficacy to include other practical aspects of achievability such as distance of the venue. In order to name each theme, we reflected on the central organizing concept of each theme-we wanted the name of each theme to provide an insight into a potential explanation for the findings of phase one. To facilitate the process of writing-up our findings, we identified suitable participant quotes that were representative of the different facets of each theme.
Pseudonyms have been used throughout the manuscript to protect interviewees' identity. When introducing participant quotes, to allow the reader to make contextual inferences, we provide the interviewee's age (e.g., 53), whether they belonged to the top or bottom 10 percentile on the measure of intentions (e.g., top 10%), and whether they were currently participating in sports (i.e., current sport participant), had previously participated in sports (i.e., past sport participant), or had no prior experience participating in sports (i.e., no sport history) in parentheses.

| Methodological integrity
We sought to ensure methodological integrity of the qualitative data by maintaining fidelity and utility. 34 Fidelity was maintained using groundedness, wherein direct quotes from participants are presented alongside our interpretations of the findings, thereby depicting their original account of the experience. 34 In addition, we interviewed a diverse range of individuals (i.e., those who scored within the top and bottom 10 percentiles on the measure of intentions). Seeking multiple perspectives allowed for a deeper understanding of why certain predictors may be relevant to the intentions of adults over 35 years to participate in walking sports. 34 Utility was preserved through contextualization, whereby contextual features of interviewees (where relevant) are described to indicate how they may be associated with inter-individual variations in perspectives. 34

| Descriptive statistics and preliminary analysis
The data were normally distributed (i.e., skewness and kurtosis values for the variables under consideration were within the range of −1 and 1), and hence, analyses were conducted with raw data. For categorical variables, when descriptive analysis showed that very few participants chose a specified response option, we combined some groups to form more evenly split categories. For education level, secondary, primary and pre-primary levels were combined. For relationship status, divorced, widowed, and separated were combined. For income, categories below $20 800 were combined. For primary mode of transport, bicycle was combined with the pre-existing "personal vehicle" category, while rideshare, bus, train, and ferry were combined into a new category called "shared transport." For ethnicity, limited responses from "non-Caucasian" ethnicities meant that these were all combined into one category.
Means and standard deviations for the variables studied, along with skewness, kurtosis, and their correlations with intentions are presented in Tables 1 and 2. Bivariate correlation analyses indicated that ethnicity and all of the tested psychosocial and program-related predictors were significantly associated with intentions to participate in walking sports.

| Social-ecological predictors of intention
A four-step hierarchical multiple linear regression was conducted to examine the association between various personal, psychosocial, program-related, and environmental predictors and intentions to participate in walking sports, as presented in Table 4. The results revealed that, at step one, personal predictors accounted for 7.2% of variation in intentions-this was non-significant, F (11, 270) = 1.915, p = 0.037, with a small effect size (Cohen's f 2 = 0.08). At step two, the addition of psychosocial predictors explained 41.6% of the variation in intentions, and this change in R 2 was significant, F (4, 266) = 54.097, p < 0.001, with a large effect size (Cohen's f 2 = 0.71). At step three, the inclusion of program-related predictors explained 6.2% of the variation in intentions, and this change in R 2 was significant, F (11, 255) = 3.218, p < 0.001, with a small effect size (Cohen's f 2 = 0.07). Finally, at step four, the incorporation of environmental predictors explained 1.1% of the variation in intentions, and this change in R 2 was nonsignificant, F (2, 253) = 3.256, p = 0.040, with a small effect size (Cohen's f 2 = 0.01). Together, the full model explained 56.2% of the variation in intentions to participate in walking sports, (F (28, 253) = 11.600, p < 0.001) with a very large overall effect size (Cohen's f 2 = 1.28). In the full regression model, perceived health status (β = 0.139, p = 0.010), attitude (β = 0.324, p < 0.001), subjective norms (β = 0.213, p < 0.001), and the maximum distance respondents were willing to travel (β = 0.150, p = 0.005) were found to be significant predictors of intentions.

| Qualitative findings
We identified four key themes relating to interviewees' intentions to participate in walking sports: "What's in a name? A need for attractive nomenclature," "Overcoming a fear of the unknown," "Congruence between personal physical activity needs and sport offering," and "Promoting achievability of the activity." The first two themes relate to the novelty of walking sports, whereas the second two themes relate to individual differences in the appeal of walking sports.

| What's in a name? A need for attractive nomenclature
Participants highlighted the lack of visibility of walking sports. None of the interviewees had ever heard of walking sports prior to research participation. Consequently, they noted that their initial impression, based on a description of the activity, was that it "sounded lame" and uninteresting. Anthony (age 48, bottom 10%, no sport history) said: Playing just a standard or existing sport, but just at a walking pace, I find it a little bit T A B L E 4 Summary of hierarchical multiple linear regression analysis predicting intentions to participate in walking sports. embarrassing…if I was to play something like that, it would probably be quite good for me and I imagine that if I was able to get past that hurdle, I would probably have a bit of fun, but yeah, I don't know…in the back of your mind, you know you're meant to run, but this is what we're doing instead.
To address this barrier, participants offered suggestions, rather than emphasizing what is absent from the activity, positive descriptions may foster favorable attitudes toward walking sports, and hence, increase the likelihood of individuals participating. Joanna (age 58, top 10%, no sport history) pondered: I found it…incredibly appealing. I just think it's about, probably not focusing on the fact that it's walking, but focusing on the other things that it brings with it. So, I think just even calling it walking sport probably makes it sound very unexciting. But I mean, it describes what it is, but it doesn't make it sound appealing generally, but you know it's attainable, it's helpful, beneficial, it's fun, it's social. I think, if you look at using other adjectives around that, then that would make it more appealing.
Furthermore, when walking sports were described as skill-focused sports, rather than traditional sports without running, interviewees seemed to hold more favorable opinions of walking sports. Nathan (age 38, bottom 10%, no sport history) said: I hadn't thought of that skills aspect, I think the more I think about that, I actually really like that, and that's probably almost the main appeal to me now that I think about it. Even if you're learning a sport for the first time… you can learn it at a walking pace without some of the other things you need to learn along with it…I think it would be great, yeah.
Participants highlighted that, given the novelty of the activity, it was challenging to visualize what the activity would look like in action-further insight would be beneficial to build an understanding of, and potentially improve their attitude toward, the activity. Isabel (age 42, top 10%, current sport participant) said: One of the things that came immediately to mind when I was trying to picture it was "Oh my God, how do they cover that much distance?"

T A B L E 4 (Continued)
And that's a really hard thing to imagine…the ball moves way faster than I can walk…so, [am I] going to be just standing around waiting for people to walk up and down before the next thing happens?…They're the sorts of things that I would be interested in just being able to see a new sport like that, or a variation of a sport like that, in action. I think it's important that we get to see it so that we can picture ourselves in it…Maybe if I saw what it was, then I'd be able to give it a fair judgment as to whether I can see myself doing it, or whether that's something that maybe isn't really for me.
Therefore, the perception of walking sports among the general public may rely to a large extent on its visibility and the nomenclature attached to such a novel form of sport, and the way in which these sports are described and advertised.

| Overcoming a fear of the unknown
Another notable participation barrier linked with the novelty of walking sports was the lack of familiarityinterviewees were hesitant to participate alone in an activity where they were unsure of what to expect. Martin (age 40, bottom 10%, current sport participant) said: I think it's just getting out to do something different and go somewhere where you don't know anyone and you haven't played the exact sport before. I guess that would probably be the biggest [barrier] to be honest…I think it's just the you know, the motivation to do something new.
Corresponding with the significance of subjective norms in predicting intentions to participate in a walking sport, as suggested in the quantitative findings, interviewees identified the critical role played by friends and family in initiating participation. Martin went on to add: I think if you went along with a friend, you'd be much more inclined to go. I think you'd be more likely to stay just because the comfort level would be there already because you're there with someone that's familiar. So yeah, I definitely think that would help…I probably wouldn't go alone.
Furthermore, highlighting the facilitative role of family, Joanna (age 58, top 10%, no sport history) mentioned how she would like to accompany her husband to participate in walking sports to reduce the social barrier. By doing so, she hoped to improve his level of physical activity: I come from quite an active background, [my husband] doesn't, so that's why I'm looking for something that will encourage him. Because while he recognizes it's important, it's not something he's ever done. It would be nice to help with that process, because often people find it more difficult to try something out alone or by themselves.
Once this initial hurdle was crossed, team membership may be mobilized to sustain participation. Maya (age 41, top 10%, past sport participant) said: That's where you start to forge friendships, you know. It becomes bonding through positive reinforcement. And sometimes it feels a bit more like you're accountable to someone else then. If you've told someone you are going to be there next week, and if you don't show up, you almost feel guilty.
These perspectives depict the importance of subjective norms, and support from loved ones, in enabling participation in physical activities such as walking sports.

| Congruence between personal
physical activity needs and sport offering Interviewees described their intentions to participate in walking sports in relation to the fit between their physical activity needs and what walking sport has to offer. Specifically, individuals were more inclined to consider participation in a walking sport if it satisfied what they were looking for from a physical activity. In line with our findings in phase one, individuals holding favorable attitudes toward walking sports expressed high intentions to participate in walking sports. For individuals drawn toward walking sports, the low-impact nature was particularly attractive-this adaptation provided an opportunity to participate in team sports within a safe environment. Maya (age 41, top 10%, past sport participant) said: It sounds like something I might be interested in. With netball, if you run and turn suddenly and then you stop, it's not very good for your knees, but a walking version of netball is going to take that away…In that sense, I guess it sounds pretty alright to me.
Contrastingly, to those less-inclined to participate, the low impact was especially unappealing-these individuals were typically extremely active. Charles (age 37, bottom 10%, current sport participant) said: Being a person who is still reasonably fit, I just like the challenge of maybe pushing yourself a little bit-that's attractive for me. So, taking that high impact out sort of changes things for me.
We identified two key participant-related predictors pertinent to this congruence-age and prior sport experience. Typically, interviewees who expressed high intentions to participate in walking sports were relatively older in age (over 50 years of age), moderately active, and had previously played sport. Heidi (age 52, top 10%, past sport participant), who had played sport at an elite level before having to retire, said: As you get older, you can't really perform, and a lot of people, like me, would still like to get out there and [play sport], so it would be awesome to do this sort of stuff when you can't [play sport] at an elite level anymore.
Specifically, for those who had played team sport, the idea of team membership was particularly appealing. Matilda (age 53, bottom 10%, current sport participant) said: I've always loved team sports; I like the teamwork. I like being part of a team, and being able to play the sport that maybe you did when you were a kid. The same but just walking.
In contrast, those expressing low intentions to participate in walking sports either had limited sport experience and/ or were satisfied with their current level of physical activity. Lydia (age 55, bottom 10%, no sport history) said: I don't think that I need anything more than my walking and my fitness classes because I can do that quite easily. I don't have to meet up with people and follow someone else's timetable…I just like to do whatever I want to do.
Nonetheless, those who were less inclined to participate in walking sports expressed positive attitudes in that they saw value in the activity for others, and potentially even for themselves in the future. Ryan (age 55, bottom 10%, current sport participant) said: At this point in my life, probably not, but I think in the future definitely, as I start to slow down. I'm 56 this year, I'm starting to slow down a little bit so I can see myself in the next five years probably taking up something like that, as I start to slow down, definitely in the next 5 to 10 years I might [consider playing a walking sport].
Therefore, the fit of the activity with each individual's unique needs and circumstances may determine to whom walking sport appeals.

| Promoting achievability of the activity
Those inclined toward walking sport participation revealed having typically low self-efficacy in relation to physical activity, but a desire to be more physically active. We identified two participant-related predictors that may be relevant to self-efficacy-health and perceived physical capability. Health was noted as having a bidirectional link with intentions to participate in walking sports. For those inclined to participate in walking sports, improving health and well-being was a key motive. For those expressing health concerns, the walking element made walking sports a particularly appealing option-while these individuals expressed low self-efficacy for traditional sport participation, they felt that the demands of a walking sport were of an "achievable" level of activity. Joanna (age 58, top 10%, no sport history), who had a diagnosis of fibromyalgia, said: Given my age and my fitness level…I have to be conscious of my energy at times. I had considered doing team sports, but to go and play something like netball or soccer or hockey, 90 minutes of running is no longer possible for me. I'm at the point where I'm not going to get to that level again in my lifetime, but to be able to do it at a walking pace is something quite different and very achievable…so it just ticked all my boxes really.
Equally, it was noted that an individual's perceived physical capabilities would be pertinent to self-efficacy for walking sports. Heidi (age 52, top 10%, past sport participant) said: [I would] probably [try it out], I just don't know, physically, whether I am capable of it. Like I said, I'm not very active anymore, so I don't run every day, I just walk the dogs, so it would be whether I'm feeling that way in my mind, whether I'm thinking I can't do this, or I can, so it's probably more mental things [that would stop me].
To some degree, it was noted that interviewees' perceived physical capability relied on a supportive, social, participation environment. Isabel (age 42, top 10%, current sport participant) said: I totally would, but I would also be conscious of the fact that I don't already know how to play [a sport]. I feel like if walking sport is a place where you don't feel like you're letting the team down by being no good at it, or having never done it before, that would make me come back, having this sort of nonthreatening supportive group of people.
However, in relation to physical distance, close venue proximity was preferred. Interviewees suggested that they would be especially encouraged to participate in walking sports if sessions were conducted at a walking distance away from their residence. Maya (age 41, top 10%, past sport participant) said: If you're going to do walking sports, [you] have to have a sense of walkability…make it really localized so it's easy for you to just chuck on a pair of shoes and some gym gear and walk down to your local park and play.
As such, achievability may be of relevance to walking sport participation. Hence, finding ways to promote achievability, whether in terms of self-efficacy or localized venues, for walking sports may encourage participation in such programs.

| DISCUSSION
In this study, we adopted a two-step mixed-methods approach aimed at identifying personal, psychosocial, program-related, and environmental predictors of intentions of adults over 35 years to participate in walking sports, using the social-ecological model. From our statistical analysis, we identified four (one personal, two psychosocial, and one program-related) significant predictors-perceived health status, attitudes toward walking sport, subjective norms, and the maximum distance respondents were willing to travel to participate. Subsequently, from our thematic analysis, we also identified that the novelty of walking sports and individual differences in the appeal of physical activity were potential determinants of intention to participate in walking sports.

| Personal predictors
At the personal level, we found in the survey perceived health status to significantly predict intentions of adults over 35 years to participate in walking sports. This is consistent with existing literature exploring adults' physical activity through a social-ecological lens that suggests health perceptions and functional capacity as key personal predictors. 16,30 In contrast, aging adults reporting poor physical health and ability were less inclined to participate in sports-including walking sports. 9,14 Paradoxically, although sport participation improves physical health outcomes, it is typically healthy adults who see themselves as capable of participating in sport, and ultimately participate in sport. 9 The importance of perceived health status was reinforced in our qualitative findings, specifically in relation to the individual differences in appeal of walking sports. In terms of congruence, such opportunities may be especially attractive for those who wish to remain active, enabling these individuals to positively negotiate the aging process through sport participation. 8 As highlighted by interviewees in our study, and supported by prior literature, self-efficacy may be important for walking sport participation, but also for physical activity more broadly. 14,30 Some aging adults expressed frustration that their poor health restricted their ability to participate in sport, especially if they had been fairly active in the past. 16 Self-efficacy can be further reinforced through the social environmentaccepting all ability levels has been attributed as an important influence in maintaining long-term participation. 15 Therefore, given the reduced physical demands for participation (i.e., walking pace, low-impact, non-contact sport), walking sports may be uniquely positioned to foster self-efficacy for aging adults who may, for any reason, feel incapable of participating in traditional sports, but wish to be involved in sport. However, given the heterogeneity of this demographic group, and the importance of promoting achievability, walking sport programs should cater to participants' abilities and needs. Supporting skill development and enjoyment may be vital to the experience of self-efficacy and long-term engagement of aging adults in walking sports.
In other studies, prior sport experience, physical activity, ethnicity, socioeconomic status, education, employment, and relationship status were noted as other relevant personal predictors of aging adults' physical activity engagement. 14,17,18 Elsewhere, differences in receptiveness to physical activity were noted based on gender and age-women and older adults were less inclined toward long-term physical activity participation. 30 Our findings did not replicate these results. Such differences may result from the fact that our study assessed intentions to participate in a hypothetical sport offering rather than an actual offering, or one that participants might have previously experienced.

| Psychosocial predictors
At the psychosocial level, attitudes toward walking sports and subjective norms were significant predictors of intentions of adults over 35 years to participate in walking sports. These findings are in line with the theory of planned behavior, which emphasizes the relevance of both attitudes and subjective norms in predicting intentions. 19 Individuals holding positive views of physical activity and beliefs regarding health benefits of participation have been found to be more inclined to be physically active. 16,17 In line with existing research, and supported in our qualitative findings we noted that individuals who were initially attracted toward a walking sport typically had a history of physical activity and sport participation-these individuals tended to value the health benefits of participation and held strong sport identities. 14 Also, interviewees noted that the lack of awareness and visibility of walking sports meant that they held unfavorable attitudes toward walking sports simply based on the name. The importance of awareness of walking sport has been previously noted as vital to initiating interest; correspondingly, a lack of awareness impeded or delayed participation. 14 Cholerton et al. 14 also noted that those who engaged in walking sport enjoyed the skill acquisition and tactical elements of the sport. Therefore, as highlighted in our findings, to attract new participants, it may be worth emphasizing aspects of what a walking sport offers (e.g., skill acquisition) while also noting what is absent (i.e., running).
Interviewees also reinforced our statistical findings regarding the importance of subjective norms. Irrespective of prior (traditional) sport experience, given the novelty of walking sports, some individuals expressed social anxiety, apprehension, and uncertainty regarding what to expect from that type of sport activity. Prior research also indicates that openness to experience is inversely related to age, supporting the novelty avoidance discussed by these individuals. 35 Having to attend alone and socialize with a group of strangers was a major physical activity barrierattending with a peer and encouragement from family have helped address this barrier in previous studies. 15,16 Furthermore, it may be especially important to ensure a welcoming and supportive environment in such activities, and harness support from loved ones, to encourage engagement from those who may be socially inhibited, hesitant, or contemplating sustained participation. In Western cultures, participation in sport is highly regarded for health and other benefits-this contributes to positive norms surrounding sport participation. 36 Nonetheless, age-related stereotypes regarding appropriate sport participation are widespread, and a recent systematic review indicated that addressing age-related stereotypes was associated with improved physical activity and self-perceptions of aging. 37 Therefore, reducing the salience of age-related stereotypes may foster positive subjective norms, thereby encouraging sport participation for this demographic group.
Social support and perceived behavioral control were not identified as significant predictors of intentions to participate in walking sports. Our findings for social support were largely in line with that of a meta-analysis that indicated high variability in effect sizes and inconclusive associations between social support and physical activity. 38 These findings suggested that the timing and context of social support may be important to consider-social support, when perceived as coercion or pressure, may have negative implications for physical activity. 38 Nonetheless, although prior research indicates that perceived behavioral control may be a strong predictor of adults' physical activity intentions, 39 our findings did not support such associations. Differences may be a result of the fact that we focused on adults aged over 35 years, and as such, the mean age of participants in our study (M age = 46.08) was significantly lower than of prior research that focused primarily on older adults (M age = 72.20). 39 Perceived behavioral control for walking sports may be particularly important to older adults, who may be especially vulnerable to reduced functional capacity, to a greater extent than younger and middle-aged adults.

| Program-related and environmental predictors
At the program level, the maximum distance respondents were willing to travel was a significant predictor of intentions to participate in walking sports. This suggests that individuals inclined to participate in walking sports were willing to travel a longer distance to access a walking sport program. This contradicts existing literature where aging adults expressed a preference for venues that are in close proximity to their home and could be accessed easily. 16 Our qualitative findings aligned with past researchalthough interviewees expressed a preference for venues in close proximity. Taken together, our findings indicate that, although close proximity may be ideal, those inclined to participate in walking sports, may be willing to travel a reasonable distance to access a program. Our findings may be explained by the fact that an overwhelming majority of respondents reported having their own personal vehicle rather than relying on shared transportation, which may permit increased accessibility of distant venues.
None of the other program-related predictors were identified as relevant to intentions to participation in walking sports. Aging adults may be a heterogeneous group, having diverse preferences regarding programspecific features, and hence, requiring a variety of strategies to encourage physical activity engagement. 18,30 Notably, despite existing literature highlighting cost as key participation barrier, 8 this was not supported in our present findings. This may be because approximately twothirds of respondents earned above the median personal income in the country. Finally, socio-economic (dis)advantage and neighborhood environment were not found to be relevant predictors of intentions to participate in walking sports. This finding is in contrast to a review that suggests that neighborhood safety, including the presence of appropriate footpaths, was positively associated with physical activity. 40 These differences may be attributed to the large proportion of individuals in our sample living in areas rated in the upper half of the IRSAD-features of the neighborhood such as crime, safety, and appropriate facilities may be irrelevant to those living in wealthier neighborhoods.

| Strengths and limitations
In term of strengths, our research is the first mixedmethods investigation of the predictors of walking sport participation for adults over 35 years. We highlighted the significant predictors of walking sport participation, and gained insight into why these predictors may be contextually relevant to the target group. Furthermore, while previous investigations have primarily focused on personal and psychosocial predictors, the social-ecological approach permitted us to simultaneously examine predictors of intentions to participate in walking sport at various (i.e., personal, psychosocial, program-related, and environmental) levels of the social-ecological model. Finally, to complement existing research pertaining to the experiences of walking sport participants, another original contribution of our study is that we considered the perspective of the general population who fit the target population (in terms of age) but had no prior experience with walking sport participation. This allowed us to consider diverse perspectives regarding the extent to which walking sport is appealing to a wider range of potential participants. This information is especially relevant to support the growth of walking sport programs in ways that are acceptable and sustainable in the long run.
While interpreting the findings of the present study, there are some limitations that need to be considered. The quantitative data were obtained using a cross-sectional self-report online questionnaire, limiting our ability to make causal inferences. Furthermore, perhaps as a result of utilizing an online survey method, the mean age of respondents was on the lower end of our inclusion criteria of 35 years and older (M age = 46.08, SD = 9.75). Therefore, we are unsure as to the extent to which our findings apply to those at the higher end of the age spectrum. Future research may use additional methods to target older adults. It is important to emphasize that our findings may not adequately represent views from less educated, lower socioeconomic, and non-Caucasian ethnic groups. Future research should endeavor to specifically explore the views of such underserved groups to design and implement physical activity initiatives in line with their specific needs and values. Also, worth noting is that the physical activity levels of our sample were generally higher than, and may not be representative of, the general population. It may be that those who were more active were more inclined to participate in a study focused on sport participation. Nonetheless, our findings are generally in line with existing literature, including studies with older age groups. 8,14,16 Finally, our findings relate to the predictors of intentions to participate in walking sports; therefore, we are unable to examine predictors of actual walking sport participation (i.e., behavior). Future research may wish to investigate predictors of actual walking sport participation.

| Applied implications
Our findings may carry practical utility in terms of the appeal, feasibility, and sustainability of communitybased walking sport programs. First, in terms of program recruitment and advertising, interviewees highlighted two key perspectives. The first was the lack of visibility of walking sports, which precluded participation. Once the visibility hurdle was crossed, the second deterrent was the negative framing of the activity. Age-related stereotypes have notable implications for physical activity engagement. 37 Therefore, it may be beneficial to reframe the description of walking sports as being skillfocused rather than lacking cardiovascular exertion. This may reduce the salience of age-related stereotypes (and improve attitudes and subjective norms) in relation to walking sports, thereby increasing its appeal to adults over 35 years.
Second, in terms of feasibility, in addition to fostering positive attitudes (as suggested above), it would be necessary to target segments of the population who may be more inclined to participate in walking sports.
Specifically, it may be advisable to focus on individuals who are seeking a low-impact team-based physical activity option. Further to this, accentuating the achievability of activity involved, as well as the team membership and social element of such programs (i.e., a welcoming and non-judgmental environment), may be advantageous in attracting new participants who may be otherwise less inclined to participate. 12 Finally, in terms of sustainability, the program environment may play a vital role. Given that interviewees highlighted the importance of perceiving self-efficacy for walking sports, in line with social learning theory, it may be imperative for participants to experience success (i.e., performance accomplishments) early on to support sustained participation. 41 Correspondingly, it may be worthwhile for organizers to cultivate a supportive environment that permits participants to experience self-efficacy. In line with existing physical activity research, promoting a motivationally supportive environment that permits participants to experience autonomy, competence and relatedness, may reinforce the positive experiences and long-term participation within such programs. 42

| CONCLUSION
Our research, adopting a mixed-methods approach, has investigated intentions of adults over 35 years to participate in walking sports. Using quantitative data, we identified perceived health status, attitudes toward walking sport, subjective norms, and the maximum distance respondents were willing to travel to participate as key predictors of intentions. Using qualitative data, we found that the novelty of walking sports and individual differences in the appeal of physical activity may explain why these predictors were important. Overall, our findings support the utility of the social-ecological approach in predicting intentions of adults over 35 years to participate in walking sports. Together, these findings offer insight into the complex interplay of personal, psychosocial, program-related, and environmental predictors of intentions of adults over 35 years to participate in walking sports. Addressing these elements of a walking sport program would make such programs more appealing to potential participants, and ultimately, more feasible and sustainable to conduct in the long run. Policymakers may wish to consider these findings to promote physical activity and health in aging adults.

| PERSPECTIVE
Given the large proportion of insufficiently active adults globally, and that physical inactivity is associated with a wide range of chronic diseases, it is important to identify acceptable and feasible opportunities to engage adults in physical activity. Walking sports have gained popularity as a potential means to engage adults in sport and promote physical activity behaviors; however, there is limited evidence regarding appeal and feasibility to support its implementation and delivery. Our research highlights the multitude and complex interplay of factors from the social-ecological model that predict aging adults' intentions to participate in walking sports. Consideration of these diverse predictors may be imperative to engaging new participants (including insufficiently active adults) in walking sports (and physical activity behaviors more broadly). Broadly, these findings draw attention to the idea that, when attempting to engage adults in a specific physical activity, there is a need to direct our focus beyond modifying individual-level factors, to address contextual/ structural factors that are relevant predictors of adults' intentions to participate in that physical activity.

FUNDING INFORMATION
Hamsini Sivaramakrishnan is a PhD student at Curtin University supported by a Curtin International Postgraduate Research Scholarship (CIPRS).