What Do We Know about the Use of the Walk-along Method to Identify the Perceived Neighborhood Environment Correlates of Walking Activity in Healthy Older Adults: Methodological Considerations Related to Data Collection—A Systematic Review

: Background: The “walk-along interview” (WAI) is a qualitative spatial method that consists of a researcher walking alongside a participant during the time of an interview to identify perceived neighborhood environments. The use of the WAI method increased in various disciplines, including the ﬁelds of public health and gerontology, to assess the relationship between the individual, spaces, and walking activity. However, how and in what settings the WAI method has been implemented with healthy older adults needs to be documented and synthesized. Objective: Our aim is to conduct a systematic review of published studies that have used the WAI method to identify the perceived neighborhood environment correlates of walking activity in healthy older adults, with a speciﬁc focus on the methodological aspects related to the data collection of this method. Methods: Following the PRISMA guidelines, PubMed, Web of Science, Scopus, and SocINDEX databases were systematically searched with no limitations on publication date. Results: From 99 articles identiﬁed, 31 met all inclusion criteria, totalizing 1207 participants. Description of the method through the assessment of participants and environmental characteristics and the data collection (before, during, and after WAI characteristics). Conclusions: This review provides detailed information WAI method to assess perceived neighborhood and walk activity among healthy older adults. WAI provides different sets of opportunities and challenges. Some suggestions, such as exhaustive participants’ socio-demographics, anthropometric descriptions and data collection methods, were highlighted to be essential elements when conducting WAIs. In addition, the current ﬁndings of this review could serve as a basis for researchers, students, and the professional community who wish to apply the WAI.


Introduction
According to the United Nations World Population Prospects [1], by 2050, one in six people in the world (16%) will be over the age of 65, and one in four people in Europe and North America (25%) could be 65 or older. Numerous countries are already facing the unprecedented challenges posed by these anticipated changes, and new tools and solutions must be explored today to better meet the future needs and expectations of older adults [2].
Then, how do we gain accurate and reliable information from qualitative spatial studies on walking elsewise by walking? The present work was specifically designed

Data Extraction
Data extraction was performed by two independent reviewers (VA and MC) following a prebuilt table including information about: (1) study characteristics; (2) sample description; (3) data collection method. The two reviewers (VA and MC) compared the data for consistency. Any discrepancies between these two reviewers were resolved at a consensus meeting. If disagreement persisted, a third reviewer (NV) was consulted to reach a final decision. Figure 1 shows the flow diagram of the articles identified by the literature search, screened for eligibility, and included in the systematic review. A total of 164 articles were identified through database searching, 47 identified through Web of Science, 84 through Scopus, 18 through PubMed, and 15 through SocINDEX. Additional article (n = 1) was identified through hand searching. After removing duplicates (n = 65), 99 records were identified. After screening titles and abstracts, 37 full texts were reviewed for eligibility, of which 31 articles fulfilled the inclusion criteria and were included in this systematic review. There is no question that systematic review quality is highly dependent on the Sustainability 2022, 14, 11792 4 of 37 literature search(es) used to identify relevant studies. It, therefore, follows that searching one or more electronic databases is often supplemented by searching by hand (the "manual search approach" or "hand searching") to identify additional primary studies for systematic reviews. As such, hand searching is recognized a critical part of the review to find materials not found through traditional searches, e.g., [39,40], and is now a common process in systematic reviews articles, e.g., [14,41,42]. Indeed, even when these materials are included in electronic databases, they may not contain relevant search terms in the titles or abstracts or be indexed with terms that allow them to be automatically identified as eligible studies for the systematic review. This was precisely the case for the study from Yoo and Kim [40] that was included in our systematic review through hand searching.

Study Selection
identified. After screening titles and abstracts, 37 full texts were reviewed f of which 31 articles fulfilled the inclusion criteria and were included in th review. There is no question that systematic review quality is highly depen literature search(es) used to identify relevant studies. It, therefore, follows th one or more electronic databases is often supplemented by searching by han ual search approach" or "hand searching") to identify additional primary stu tematic reviews. As such, hand searching is recognized a critical part of the re materials not found through traditional searches, e.g., [39,40], and is now a c cess in systematic reviews articles, e.g., [14,41,42]. Indeed, even when these included in electronic databases, they may not contain relevant search term or abstracts or be indexed with terms that allow them to be automatically eligible studies for the systematic review. This was precisely the case for the Yoo and Kim [40] that was included in our systematic review through hand  Table 1 summarizes the characteristics of the 31 included articles from cent to the oldest publications.

Study Characteristics
Publication year. Figure 2 illustrates the number of included articles p year. The publication year of the 31 included studies ranged from 2012 [5] to study (3%) was published in 2012 [5], 1 (3%) in 2014 [41], 1 (3%) in 2015 [42  Table 1 summarizes the characteristics of the 31 included articles from the most recent to the oldest publications.
Disability and Rehabilitation Not mentioned.
Van Cauwenberg et al. [5] 2012 Belgium Environmental factors influencing older adults' walking for transportation: a study using walk-along interviews.
International Journal of Behavioral Nutrition and Physical Activity Not mentioned.
This table shows the basic information about the 31 included articles in the systematic review: author names, year of publication, country of the first authors' affiliation, article's title, journal where the article was published, and the funding sources.

Participant Characteristics
Basic socio-demographic information of the 31 included articles is reported in Table 2, and the exclusion/inclusion criteria for the participants of each included study are reported in Table 3.
Sample size. Among the 31 included articles, 1 (3%) did not report the number of healthy older participants [53]. For the other 30 articles (97%), a total sample of 1207 healthy older adults participated to the WAI. The mean sample size was hence n = 40.2 (±44.9), ranging from n = 4 [35] to n = 173 [46,47] healthy older participants.
Setting. All of the included studies (n = 31, 100%) were conducted in an urban setting. Of these, one (3%) study addressed urban and semi-urban settings [5], and one (3%) urban and rural settings [58]. No studies were conducted solely in a rural setting.
Time of day. Of the 31 included studies, 5 (16%) reported the times of day the WAIs were conducted [5,45,47,56,58]. Van Cauwenberg and colleagues [5] indicated that they conducted interviews during the day, Herrmann-Lunecke and colleagues [53] indicated that interviews were conducted in the morning on weekdays. Lager and colleagues [42] reported that participants chose to walk around 10 a.m. or 2 p.m. and Sun and Lau [55] conducted the WAI spanning weekdays and weekends, from 8:00 a.m. to 6:00 p.m. Curl and colleagues [45], reported that the WAIs were conducted on days with mild temperatures (10-15 degrees Celsius), and in which there was some light rainfall.
Data collection duration. Of the 31 included studies, 14 (45%) did not report how much time has been spent on the WAI [10,23,25,28,[32][33][34][35]44,45,48,50,51,58] (see Table 4). Among the other 17 included studies (55%); data collection lasted between 5 weeks [57] and 20 months [40], with a mean of 4.8 months. Data collection lasted 2 months in 5 studies (16%) [6,8,45,55,58], 3 months in 3 studies [9,32,56], 4 months in 3 studies [5,51,57], 5 months in 2 studies [7,46], 8 months in one study [41] and 10 months in one study [54].     [P] indicates a pseudonym. [C] indicates that the value was calculated by the reviewers. WAI: walk-along interview; HKD: Hong Kong dollar; GPB: British Pounds; CNY: Yuan; RCF: residential care facilities; GCSE: general certificates of secondary education or its equivalents; BME: black and minority ethnic. The recruitment strategy intended to reach a varied sample in terms of age, gender, education level, employment status, and cultural background. However, the large geographical scale and time limitation of the study led to a convenience sampling, based on the willingness of the people we met in the organizations. Knowing the mixed use of language in the Brussels-Capital Region (most spoken: French, English, and Dutch), only participants skilled in Dutch, French, or English with a minimum age of 18 years were included.
Herrmann-Lunecke et al., 2021 [53] Participants were required to speak Spanish and to have lived in the neighborhood for at least the last two years.
Sun and Lau 2021 [55] Living in the study areas and familiar with the designated routes, aged 65 or above, and can walk and use public transport without aids. Residents were eligible to participate in the larger study if they: (a) had lived in one of the neighborhoods for at least 1 year, (b) were not working or were engaged in part-time paid employment, (c) were able to converse in English, and (d) were able to access the community, either alone or with assistance.

Saint-Onge et al., 2021 [8]
Individuals were eligible to participate if they were (1) tenants of one of the three study sites; (2) able to walk four 10 min sessions, and (3) able to communicate in either French, English, or Spanish. They were excluded if they reported having an intellectual, visual, or auditory impairment that could significantly impact walking safety and ability.
Carroll et al., 2020 [50] Irrespective of their physical ability or potential impairment.
Hand 2020 [51] Selection criteria were age 65 years or more, able to participate in an interview in English, not working full-time, and able to move about the community in some way.
Cao et al., 2019 [6] This study included those aged 55-64 because that Singapore is ageing rapidly and their opinions are valuable for future developments. To be recruited, participants needed to be able to speak either English or Mandarin Chinese, living in Yuhua East, or living in the nearby neighborhood but walking to the facilities in Yuhua East on a daily or weekly basis.

Macintyre et al., 2019 [25]
Any adults aged 60 or over in 2018 were considered eligible to participate if they lived or spent a large amount of time (i.e., a minimum of one or more hours every two weeks) in Old Moat when the study occurred. Participants were excluded from the study if they had a diagnosis of dementia, since this could affect their ability to participate in the interviews. Participants who were able to complete walk-along interviews were prioritized for recruitment, although participants who preferred to participate in a sitting down photo elicitation interview were also recruited. Inclusion criteria were broad: they had to self-identify as men age 65 or older, be able to communicate in English, live in the community, and be able to move about within and outside their homes.
Lee and Dean 2018 [28] The only selection criteria were that participants had to live in the chosen neighborhoods, Wychwood and Edenbridge-Humber Valley, and be over the age of 65 years. Older adults (65 years and upward), resident of a low-or high-deprivation area, able to walk, independent in their daily life activities, and mentally healthy were eligible to participate in this research. Ability to speak English was not an eligibility criteria.
Gardner 2014 [41] Over the age of 75, living alone (as most older adults in this age category live alone) and having resided in the study neighborhood for a minimum of three years.
Van Cauwenberg et al., 2012 [5] To be included, participant had to be over 65 years old, dwelling in the community, and able to walk independently for at least 30 min.   Participant recruitment. Of the 31 included articles, the most frequently used recruitment technique (n = 16, 50%) was through visits to local organizations and/or groups of older adults (card club, local activity center, computer courses for older adults, seniors' residence, others), and information to older adults about the participation process and/or the project through information material (advertisements, letters, posters) and/or information sessions [8,9,25,28,32,34,[41][42][43]46,47,50,52,54,57,58]. In one study (3%) [53], recruitment was performed through bulletin boards and public posters placed in stores and/or street walls in the study area. In 4 studies (13%) [6,7,54,56], direct interception by the interviewer or research team was conducted in the study areas and/or facilities and open spaces. Recruitment through a sub-sample of participants from a previous study was mentioned in 4 studies (13%) [36,38,46,47]. In 6 studies (19%), no information was provided [5,10,23,43,53,55].
Compensation for participants. Of the 31 included studies, 4 (13%) studies [9,23,38,46], reported stated that a reward for their time and participation was given to participants. In 3 of the 4 studies mentioned, a $10 gift card [35], $20 gift card [44], and $25 gift card [23] were given. Močnik and colleagues [9] only noted that participants received a small token of appreciation for their participation.

How?
• Before walk-along interview. Preparing every aspects of the WAI: Interviews and/or questionnaires. Of the 31 included articles, 21 (68%) conducted an interview and/or questionnaire before conducting the WAI [5,[7][8][9][10]23,28,32,34,35,[42][43][44][45][46][47]49,51,52,54,58]. The purpose of these preliminary interviews was to gather background information about the participants, learn about their views of the site, and/or how to get a general sense of the place where the WAI would be conducted. One (3%) study [33], reported that a questionnaire was conducted with the participants, however, the authors did not mention whether it was before or after the WAI.
Interviewer training. Of the 31 included studies, 5 (16%) reported that the interviewers/researchers were trained or had training in conducting the WAI [5,7,36,57,59]. In the study of Grove [54], the researcher had training and experience in interviewing techniques and participatory research. Two studies (6%) reported that pilot WAIs were conducted in a different study area [48] or in the same study area [55].
Choice of the route. Route choice can be classified according to the typology of WAI presented by Evans and Jones [19], WAI ranges from "natural walks" to "guided walks"; the first type of WAI refers to studies in which the researcher walks a route determined by the participant; the second refers to an interview in which the route is determined by the researcher [19].
Route selection criteria. In 20 (65%) of the included studies, the criteria for selecting the route or destinations for the WAI were described: 1.
Walk in a park (n = 1, 3%) [7]. No information was provided about who chose the route; 4.
One (3%) study [54] indicated that some participants were accompanied by a family member or caregiver at the time of the WAI, however this study did not indicate whether the companion participated in the WAI.
Distance of the walk-along interview. Of the 31 included studies, 5 (16%) reported the distance of walk interview routes [36,47,52,56,59]. Sun and Lau [55] reported that the distance traveled for each WAI ranged from 443 to 635 m. Carroll and colleagues [50] indicated that it ranged from 0.4 to 4.6 km. Curl and colleagues [45] indicated that the mean walk length was 0.82 miles (SD = 0.45), Kou and colleagues [33] from 1.45 to 3.70 km, and finally Grove [54] reported that the distance traveled was 300 m.
Stops or breaks along the route. Of the 31 included studies, 4 (13%) reported whether stops or breaks were taken during the WAI [6,8,9,25]. Saint-Onge [8] mentioned that the participant was offered as many breaks as necessary during the walks, providing places where these could rest along the way, and carrying a folding stool in case the resting spaces were far away. Macintyre and colleagues [25] reported that, on many occasions, it was necessary to stop and spend some time observing the places, in order to ask more detailed questions about the participants' experiences in the places before continuing the walk. Cao and colleagues [6] reported that letting participants talk as freely as they wanted, even if the conversation went beyond the purpose of the study, allowed participants to engage in various activities along the route or stop to chat.
Of these 24 studies, 3 (10%) studies [7,36,59] reported who carried the recorder and/or video camera. In the study by Kou and colleagues [33], the participant wore the chestmounted video camera. In the study by Veitch and colleagues [7], the participant carried the voice recorder. In the study by Grove [54], the interviewer carried the audio device which was hidden during the interviews so as not to attract the participant's attention and to make the participant feel more comfortable [54].

•
After the walk-along interview: Meeting with participants. Of the 31 included articles, 2 (6%) studies [43,52] reported that the researchers met with the participants after the walk-along interviews. In the study by Luusua and colleagues [43], immediately after the guided walks, participants were also interviewed about their overall experience [43]. Lauwers and colleagues [52], reported that after all the WAI completion, all participants were invited to a workshop to thank them for their participation and to present and reflect on the results of the interim project [52].

Discussion
WAIs have proven to be a very efficient way of accessing a local community's connections to its environment [19], they are especially useful for studying spatial perceptions and practices in the physical and social environment [24]. For these reasons, a large number of works have used the WAI over the years and propose various methodological explorations of the tool [26]. However, few works offer the different methodological modalities and variants of the WAI e.g., [19,24], and to the best of our knowledge, there are no published works on the methodological modalities and variants of the WAI applied to research on aging and its relationship to the environment. Therefore, the present review sought to fill this gap from a methodological perspective of data collection.
The objective of this systematic literature review was to identify and synthesize the available evidence on the use of the WAI method to identify correlations of the perceived neighborhood environment and the walking activity of healthy older adults. As a first step, this review allowed us to describe the ways in which WAIs have been used with healthy older adults, focusing specific attention to the methodological aspects related to data collection of this method. In other words, how and in what settings the WAI method has been implemented with healthy older adults. The analysis of the included studies also revealed the strengths and limitations that the authors themselves have acknowledged in using this method, and suggested perspectives for the improvement of data collection.
In total, we found 31 published articles that fall within our selection criteria. We will discuss our findings through five main themes: 1. study characteristics, 2. participants' characteristics, 3. description of the data collection method used, 4. strengths and limits of the included studies and 5. propositions for further studies. For the analysis, we will sub-categorize in the following order: first, general information about the study selection; second, we will detail which study protocol authors refer to when using the one of the terms referring to WAI; third, we will present strengths and limits identified by the authors themselves. Finally, we will conclude with the limits of the review and perspectives when using WAI with healthy older adults.

Study Characteristics
Of the 31 included articles, about three-quarters (n = 23, 74%) were published in the last 5 years, reflecting the increasing use of this method in an older adult population. It is also worth noting that this growing interest is not limited to this target population. Indeed, a search of Web of Science databases using keywords related to WAI only (Ts = ("Marche commentée" OR "Marche commentee" OR "Parcours commentés" OR "Commented routes" OR "Routes commented" OR "Commented walk" OR "Commented Path" OR "Walk-along interview" OR "Go-along interview") yielded 194 records of which 155 (80%) have been published in the last 5 years. More largely, since the use of WAI method and other spatial qualitative methods to assess walking accessibility is increasing not only for older people, we believe that our systematic review could serve as a basis for future studies that wish to use the WAI method with older adults or another population group.
The age of participants included varied widely between those studies, from older than 55 years old [6]-our inclusion criteria-to over 90 years old [28]. Differences as well as continuities in young-old, old-old, and the oldest old adults have to be thoroughly studied, considering the long span of years covered and the significant decline in mobility and physical health mobility [56]. Included studies that tested different age groups of participants did not discuss the potential effects of age on the WAI methods used. At this point, we believe that comparing the use of the WAI method could be of interest to better appreciate potential differences in WAI implementation to help to extract relevant information for each specific age population.
Finally, basic anthropometrics characteristics (height and weight) of the participants was not reported in any included studies, so that body mass index was not reported or calculable. At point, however, it is important to mention that overweight and/or obesity is recognized to significantly affect gait stability [60,61], and to increase energy cost of walking [61]. Although these mobility difficulties are observed when walking under rather low challenging conditions, such as walking on a flat, straight, unobstructed, and short corridor in indoor environment without any distractions e.g., [64][65][66][67][68][69][70], they are exacerbated under more challenging gait conditions, including walking around curves [69], climbing stairs [70], stepping over obstacles [71,72], or walking while talking [78]. These conditions could presumably significantly increase the risk of loss of balance, trips, and falls in individuals with overweight or obesity [75,81,82]. What is worthy of note is that the above-mentioned challenging gait conditions combine walking with other physically and cognitively demanding tasks that specifically approximate WAI data collection conditions in which a researcher and participant walk together in real outdoor environments on various/irregular road surfaces while the researcher interviews the participant. With these thoughts in mind, we do believe that it is of particular interest to adapt the WAI data collection method according to the characteristics of the participants. More largely, taken together, these findings suggest that socio-demographic and anthropometric characteristics, as well as health-related fitness (i.e., cardiorespiratory endurance, muscular endurance, muscular strength, body composition, and flexibility) and the physical activity level of interviewees should be taken into account to adapt and ensure optimal WAI data collection conditions in an older adult population.

Environmental Characteristics
As the results show, all of included studies were conducted in urban settings (of these, one (3%) study addressed urban and semi-urban settings [5], and one (3%) urban and rural settings [58]), and 78% of these studies were conducted in Europe (n = 16, 52%) and North America (n = 8, 26%), therefore, the data collected in WAI and presented in this systematic review have a strong European and North American influence. It is also interesting to note that the countries where the studies were conducted are very similar to the World Health Organization's projections for the countries with the highest proportions of older people in 2050 (e.g., Canada, China, and the U.K.) (see Figures 3 and 4).

When the WAI Takes Place? Seasons, Weather, and Time of Day
The results show that 68% of the included studies reported the seasons of the periods of the year in which WAI was conducted [5][6][7][8][9][32][33][34]40,42,43,[45][46][47][48][49][52][53][54][55]57]. However, very few studies mentioned the influence of weather/temperature during data collection [9,32,35,36,45]. Močnik and colleagues [9], who conducted WAI in the winter season (1 December 2017 to 21 February 2018), indicated that due to the tropical, hot, and humid weather conditions in Singapore, they took additional measures, such as having enough time to stop and rest during the walk, as well as, reminding participants to wear comfortable clothing and footwear for the walk. Kou and colleagues [33] noted that they were unable to perform a WAI due to inclement weather. Lager and colleagues [32] noted that the WAIs were conducted in "good weather conditions [i.e., no snow, ice, or rain, and not too cold or too hot] and also noted that walks in less favorable weather conditions may have provided additional insights into the experience of older adults and, in particular, the obstacles to traversing the neighborhood space" [32]. Ambient temperature can become a factor affecting participants' willingness to walk [19]. Talking while walking and simultaneously navigating the urban environment can be challenging for people with balance disorders and visual or mobility impairments [74] without adding adverse weather conditions. However, we are aware that weather and seasonal conditions influence the walking movements of older adults [75][76][77], so if additional measures are taken when conducting WAIs, rich information could be obtained to understand the influence of weather/temperature on the relationship of the older adults and their environment. In addition to the weather, the time of day at which the WAI was conducted also has an influence since the frequency of social activity may vary depending on the place where the interview is conducted [11].      Interviews/questionnaires prior to conducting the WAI were a procedure used in a majority of studies (n = 21, 68%). Lager and Colleagues [32] indicated that the aim of the pre-WAI was to elicit experiences, feelings, and memories of the participant's daily life in the neighborhood. Saint-Onge and colleagues [8] noted that the pre-interview served to establish a relationship between the interviewer and the interviewee. Before the completion of the WAI, Grove [54] conducted a semi-structured interview together with a mapping exercise in which participants were required to identify places important to them and their usual routes, which helped to "set the stage" for the live interview that followed [54]. Furthermore, in the study of Sun and Lau [55], where the researchers were the ones who defined the route for the WAI before the WAI occurred, they provided a route map and specific information about crossings, turns, and directions to the participants [55]. The interview/questionnaire before the WAI seems to be a fruitful procedure to be able to connect with the participant, get to know his/her characteristics better, and approach the WAI based on these characteristics.

Who Is the Interviewer?
The results indicated that WAIs were mainly carried out by a researcher/interviewer (n = 22, 71%). However, very few studies (n = 7, 23%) [5,7,36,54,56,57,59] mentioned whether the interviewer/researcher was trained or if pilot tests had been conducted to carry out the WAI. We believe that this aspect is important to take into account when conducting WAIs; pilot testing is an excellent example. Kou and colleagues [33] conducted three pilot walkthrough interviews prior to the start of the study only in order to test the interview protocol and the quality of the audio and visual recording; these pilot tests were not included in the subsequent data analysis.

Interview Guide?
Although only 35% (n = 11) of the included articles reported the use of an interview guide, only 4 included it as part of the supplementary material or in the main text [5,10,36,49]. Sundenvall and Jansson [48] reported that they conducted "funnel-shaped interviews" with open questions and more specific questions, first opting for open questions, giving the opportunity to interviewees to narrate in their own words and according to their interest, then continuing with more specific questions.

Typology of WAI, Two Ways to Choose the Route
The route decision for the WAI is an important aspect to take into account. According to the typology of WAI presented by Evans and Jones [19], WAI ranges from "natural walks" to "guided walks". In 48% (n = 15) of the included studies in this review [6,[8][9][10]23,32,33,41,42,44,46,47,50,52,54], the participants set the route and thus also the distance, while in 6% (n = 2) [43,55], the route decision was made by the researcher/interviewer. In the study by Luusua and colleagues [43], the researchers did define the route for the WAI, and the WAI was conducted along a path in a park, resulting in a relatively strict interview choreography, characterized by the restriction of the participant's body movements [43]. Other dynamics for route choice were also observed. Macintyre and colleagues [25] included a plan of possible interview routes, which could be adapted according to the individual characteristics of the participants. Herrmann-Lunecke and colleagues [53] defined the routes previously with the residents in neighborhood meetings. In the study by Van Cauwenberg and colleagues [5], the participant and the researcher walked to and from the destination along two different routes, the first route, chosen by the participant and the second return route that was chosen by the researcher based on the availability of different routes. The use of two routes seems to be an interesting option when conducting the WAI, as using two different routes increases the number of different environmental stimuli encountered during the WAI and provides the participant with environments that they did not walk through [5].
At this point, it is important to note that no study has described the physical characteristics (e.g., slope, presence of sidewalks, presence of stairs, etc.) of the routes along which the WAI was conducted. We believe that this is a very important factor to consider since it is now well established that the physical characteristics of the terrain do influence the walkability of elderly people [78,79].

Duration and Distance of WAI, and Stops/Breaks?
Neither the duration of the route nor the distance of the route was precisely reported in the included studies. On the one hand, indeed, information on the distance traveled during the WAI is in 5 out of the 31 included studies (16%) [36,47,52,56,59]. Furthermore, the ranges are highly variable; for example, in the study by Carroll and colleagues [50], the distance varied between 0.4 km and 4.6 km per participant. On the other hand, the duration of the WAI was reported in 21 of the 31 included studies (68%) [5][6][7][8][9][10]23,25,[33][34][35]41,42,[45][46][47][48][50][51][52]54]. As seen for the distance, the time ranges are variable. For example, Carroll and colleagues [50] reported that the WAIs duration ranged from 30 min to more than 2 h per participant. No details about the characteristics of the interviews that lasted 30 min and those that lasted more than 2 h were reported. Cao and colleagues [6] reported that the duration of their interviews was between 9 min and over 2 h per participant. These authors observed "that shorter period of walk-along interviews can only elicit participants' limited feedback towards the surroundings, while those that take more than an hour can provide a more holistic picture of the older adults themselves and their ageing experiences, this confirms Kusenbach [24]'s conclusion that a productive time window for a go-along is about an hour to 90 min" [6] (p. 19).
Although there are two types of WAI ("nature walks" and "guided walks") [19], in which the choice of the route (which determines the distance and duration of the walk) is established by the participant or by the researcher/interviewer, we suggest that regardless of the type of WAI, the duration of the interview could be determined largely by the participant's characteristics, as well as his or her health-related fitness (i.e., cardiorespiratory endurance, muscular endurance, muscular strength, body composition, and flexibility) and physical activity level. The above-mentioned characteristics could make it difficult to propose a similar WAI duration with all participants or to comply with the "productive time window" [6,24]. For these reasons, we strongly believe that stops and/or breaks do play an important role in WAIs, especially with the older adults. While there are only 4 studies (13%) that reported having scheduled stops/or breaks [6,8,9,25] either to chat or to rest, two of them [8,25] reported that having conducted them allowed the interviewer to ask more questions and the participant to speak more freely. More largely, if we search for the meanings of WAI, the distance and duration should not only be conceived as objective aspect interacting with socio-demographic and anthropometric characteristics; the subjective aspect of them should also be taken seriously to ensure optimal WAI data collection conditions in an older adult population. 4.4.6. WAI Recording: Route Tracking, Field/Reflective Notes, and Visual/Audio Recording Even though the mapping of routes using a GPS tracking unit or manual drawing in the WAI is not very common in the studies in this review (n = 8, 26%) [6,9,32,[45][46][47]50,54], we consider that it could be very useful for subsequent analyses, to geographically identify the route and the distance traveled by each participant.
Furthermore, while field notes during the interview are a very practical tool to record in real time the events during the interview [41], field notes/reflexive notes after the interview seem to be more interesting if combined with an audio or video/photo recorder for the WAI. For instance, Sundenvalll and Jansson [48] recorded WAIs using audio recordings and photographs of the locations that participants described using during the interviews, supplemented with notes taken directly after each interview on aspects that the audio recording could not capture, such as actions. In addition, video recording of the interview seems to be an interesting tool, as it allows reconsidering the movements and places embodied during the analysis [34] and leaves a rich trace of the brief encounters with the participants in the study [43].

The Mixed-Methods Approach
The mixed-methods approach can be a great alternative for studying older people in relation to their environment as it helps to better understand the relationships between older people and their living environment [57]. Zandieh and colleagues [46] adopted a mixed-methods approach (Geographic Positioning System (GPS) technology, questionnaire, WAI), which helped to enrich the quantitative study on "perceived neighbourhood built environment attributes and outdoor walking levels" with qualitative information on "how perceived neighbourhood built environment attributes may, in the opinion of older residents, influence outdoor walking levels", which helped to triangulate the information [46].

Strengths/Limits and Propositions
WAI as a qualitative method is greatly put forward as a strength by the authors of the included studies [5,7,25,32,36,37,44,45,51,53,56,59] They underline the importance of the collection of context-specific data [33] such as the (physical, mental, social, and emotional) person-environment interplay [7,23,51,57], place-attachment [54] or micro-spaces [51]. Curl and colleagues [45] even go as far as to provide a checklist "[f]or built environment professionals, including urban designers and planners, as well as highways and civil engineers, landscape architects, municipal authorities and city centre managers, it forms an aide memoire of the elements that are critical within the outdoor environment to minimise the risk of falls by older people" (Guidance notes: Assessment of the local Outdoor Environment).
Due to time-intensive data collection, qualitative research methods are more adapted to smaller samples [6]. This raises the challenge of a representative sampling of research participants [8,9,56], and the exact recruitment terms are not specified in one-third of the studies (10/31). So, rather restrictive conditions of participation (having lived in the neighborhood for a minimum of a year (4/21) or the fact that only one language for communication allowed (7/21) reduce accessibility to the studies and foreclosed the possibility of presenting a more holistic sample, especially for the "hard-to-reach". A total of 10 out of 21 studies conditioned their participants to be aged 65 years and more. Clearly, delimiting chronological age when researching with/on/about older adults is a handy tool. However, we might miss the main challenge of the aging process, such as the transition moments and push-and-pull factors that enhance sustainable mobility patterns and walkability?
On the other hand, to the best of our knowledge, a very limited number of published studies have conducted the WAI in rural settings [49,84,85]. As an example, we can mention the recent work by King and colleagues [80], who conducted WAI, in-depth interviews, and observation to study the food experiences of older adults living in rural areas in Australia, and the work by Mackay and colleagues [81] who studied rural tourism on multigenerational family farms in New Zealand. Note that the studies by King and colleagues [80] and Mackay and colleagues [81] were not included in the present systematic review as they do not meet the inclusion criteria of the review. A relevant point to note is that all studies included in the present review (n = 31, 100%) were conducted in urban settings, while only one of these studies [58] conducted the WAI in both urban and rural settings in Ireland. In this study, Cassarino and colleagues explored the cognitive and sensory dimensions of older adults' outdoor walking preferences, in which 41% of their participants rated their place of residence as rural (village or countryside).
It is recognized that the differences between urban and rural environments are mainly related to the morphology of urban forms (i.e., characteristics of the street network, volume, and height of buildings, etc. [82]) and their functional configurations (i.e., presence of ser-vices, mix of functions [82,83], green space [84], etc.) [82][83][84]. Cleland and colleagues [83] indicated that urban-rural status is an important factor in understanding associations between environment and physical activity among mid-older aged adults. These authors have, for instance, reported that perceived levels of personal safety were significantly more favorable among rural than urban residents aged 55-65 years. Bucko and colleagues [85] reported that a number of factors differed between rural and urban environments, including access to green space or recreational facilities, neighborhood safety, availability of transportation, and presence of pavements. Jo and colleagues [84] found that the proportion of dissatisfaction with all environmental factors related to walking was higher in rural residents than urban residents aged 12-81 years; the presence of dirt or much litter was the source of dissatisfaction that most influenced overall dissatisfaction in rural areas. Cassarino and colleagues [58] observed that older adults in rural areas had lower levels of walking in their neighborhood than older adults in urban areas, which could be related to accessibility to services and amenities. These authors further reported that rural and urban environments are associated with walking for different purposes: "the choice of outdoor spaces was determined by the purpose of the walk (recreation vs. transportation) and also by the level of urbanity of the place of residence: urban dwellers walked for transport in their neighbourhood but drove to green/blue spaces for recreation, with the opposite pattern for rural dwellers" [49] (p. 7).
Aging highlights multiple challenges posed by different environments, which offer more or fewer opportunities and possibilities to their inhabitants in terms of planning and development in both urban and also rural areas [82]. In this context, it seems appropriate to apply the WAI method in every type of walking-related environment, including hence more suburban and rural areas, to have a better understanding of the experience of aging and its interaction with its environment [28,51]. While such differences might be significant, critical social sciences argue about the need to get over this dual "rural/urban" categorization, calling for more studies [86] or presenting administrative new classifications to meet such a reality as INSEE in France [87].The gender of the participants, on the other hand, was clearly reported in 26 studies, with 24 involving male and female participants (one study involved only males [35] and another one only females [51]). Nevertheless, no gender comparison was conducted, although it is now recognized that male and female individuals appeared to have different gait behavior on flat ground walking [88]. Walking time for leisure is also different between males and females and changes with age [59]. Therefore, it might be relevant for future studies to investigate potential differences between male and female individuals.
Finally, the need for additional, geographically more varied studies remains, as the setting-geographical location, urban/rural [5], season, weather and the hour of the dayhighly influences the results of the study [9,32,35,36]. This was partly identified by the authors themselves [32].

Scoping Limits and Perspectives of the Review
They are some limitations associated with the present systematic review that should be taken into consideration when discussing the findings. Limitations to this systematic review include the search strategy as only five databases were consulted. In addition, the number of articles included is relatively low (n = 31), as well as the number of participants (1102 identified as older adults included in these studies). Furthermore, articles not published in English were excluded, so it is possible that other outstanding articles may have been missed. This is unfortunately common practice; however, this could be rectified in future review articles. In addition, the fact that there are many different terms attached to the WAI method, may have limited the research results. Considering that only a limited number of studies clearly define the research protocol, we will suggest a guideline when using WAI with older healthy adults.
Research combining qualitative-participatory methods such as WAI appears to be an emerging area within the study of aging and environments, as the majority (n = 23, 74%) of the studies included in this review were published within the last 5 years. The use of WAI can provide a valuable contribution to the development of knowledge about the neighborhood/urban environment relationship and older adults as it can capture situated details about places that are not verbalized during an interview or otherwise discerned and can explore interpretations, helping to generate a deep understanding of person-place relationships [23], the strength of this particular approach lies in its assumed ability to provide access to participants' attitudes, knowledge, and perceptions regarding the surrounding environment [11]. Based on the results of this systematic review, published studies are encouraging us to provide suggestions that could help students, researchers, professionals, and communities to better plan and implement WAI with older healthy adults.

How to Prepare a "Walk-along Interview"?
Future studies should focus on the following recommendations regarding the data collection method: (1) socio-demographic (age, gender, socio-economic status), and basics anthropometrics (height, weight, BMI) characteristics assessment of the participants should be reported; (2) interviewer preparation in qualitative methods or pilot testing is necessary before beginning the study [5,7,54]; (3) conducting an interview/questionnaire before doing the WAI could represent useful and necessary procedure to have a first contact with the participant [54]; (4) consider weather conditions when conducting WAIs to take precautionary measures [9,32], as well as; (5) the season as an important factor in the analysis, as winter may be associated with less engagement of the person with the outdoor environment [89,90] and heat and sun exposure may also be factors that hinder walking [91]; (6) having an interview guide for WAI [5,8,49]; (7) and combining the two typologies of WAI (namely "nature walks" and "guided walks") [19], by selecting two different routes [5], can be an interesting technique to apply, as they can reveal more information of the participant's relationship with their environment; (8) duration and distance are two aspects that should be detailed in the studies and considered in the analyses; (9) stop/rest locations for the participant, may provide a time to ask more detailed questions about participants' experiences [25]; as well as; (10) the use of tools such as video cameras/photography in the WAI may provide valuable visual documentation of the participants' everyday life contexts and actions, and serve as excellent stimuli and reminders during data analysis [29,41]. However, we propose the WAI to be audio-taped [11,35] (lavalier microphones are the less intrusive way), (11) use reflexive notes to complete data collection [8,44]. Finally, (12), the mixedmethods approach [10,23,25,49,53], the combination of qualitative, quantitative, geospatial, or survey research methods, can provide useful insights into person-place relationships.

Conclusions
To conclude, WAI does provide different sets of opportunities and challenges. Opportunities are to implement paradigm shifts in theory to the data collection processes. This systematical review provides an overview of data collection aspects related to the use of WAI with healthy older adults. The current findings of this review add substantially to our understanding about WAIs as a main tool that will serve as a base for future studies notably for researchers, students, and the professional community.
The WAI method can also be used with other populations, e.g., autistic people [98], and should be addressed in future research, hence helping stakeholders and decision-makers to better understand the needs of the older urban dwellers, their pedestrian behavior, and their walking capacity. Finally, considering neighborhood perceptions of its inhabitants allows cities to become more inclusive and more sustainable. Funding: This study is part of the first author's Ph.D. thesis (V.A., Université Grenoble Alpes, France). This work was supported by the Caisse Nationale de Solidarité pour l'Autonomie (CNSA) in the framework of the call for projects "Blanc (session 12) 2021" of the French Institute for Public Health Research (IReSP) ("ADAM" project, grant BLANC21_254987) and by the French National Research Agency program (Early-Career Researcher (JCJC) "Citizenbench" project, grant ANR-19-CE22-0002). The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results. This work further forms part of a broader, translational, interdisciplinary research program called "GaitAlps" (N.V.).