Health & Place 71 (2021) 102660 Contents lists available at ScienceDirect Health and Place journal homepage: www.elsevier.com/locate/healthplace Active use and perceptions of parks as urban assets for physical activity: A mixed-methods study Mario Fontán-Vela a, b, Jesús Rivera-Navarro a, c, Pedro Gullón a, d, *, Julia Díez a, Isabelle Anguelovski e, Manuel Franco a, f a Universidad de Alcalá, Facultad de Medicina y Ciencias de La Salud, Departamento de Cirugía, Ciencias Médicas y Sociales, Grupo de Investigación en Epidemiología y Salud Pública, Alcalá de Henares, Madrid, Spain b Preventive Medicine Department, Infanta Leonor University Hospital, 28031, Madrid, Spain c Department of Sociology and Communication, University of Salamanca, 37007, Salamanca, Spain d School of Global, Urban and Social Studies, RMIT University, Melbourne, Australia e Catalan Institution for Research and Advanced Studies (ICREA), Universitat Autònoma de Barcelona (UAB), Institute for Environmental Sciences and Technology (ICTA), Barcelona Laboratory for Urban Environmental Justice and Sustainability, Medical Research Institute Hospital Del Mar (IMIM), 08003, Barcelona, Spain f Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, 21205, USA A R T I C L E I N F O A B S T R A C T Keywords: Parks are potential key urban assets for improved population health; however, their use is not equal among all 8 words): mixed-methods social groups. Individual and contextual factors could influence residents’ perceptions of parks and how they Parks interact with and, eventually, benefit from them. The use of complementary methodologies provides a deeper Green space understanding of the relationship between park use, physical activity (PA), and residents’ perceptions. Thus, we Physical activity Use designed a mixed-methods study to analyze differences in park use and PA, and the perceptions of parks as urban Urban environments assets for PA. We selected six parks from three neighborhoods in Madrid (Spain) with different neighborhood Health inequities socioeconomic status (NSES) for systematic social observation. We registered park users by age, PA level (low, Spain medium, and high), gender, and NSES using the System for Observing Play and Recreation in Communities (SOPARC) audit tool adapted for iOS software (iSOPARC). We also conducted 37 semi-structured interviews and 29 focus groups to analyze residents’ perceptions of parks as urban assets for PA in the same neighborhoods. We adopted a convergent-parallel design to analyze both quantitative and qualitative data, and to describe the convergence and divergence areas between them. Parks within the high-NSES were more visited, showing a higher proportion of people performing high PA (11.9%) as compared to residents of the middle (9.3%) and low- NSES (3.2%). Female visitors showed lower PA levels compared to men, especially for parks within high-NSES. The following issues were reported as influence urban park use and perceptions: park maintenance and area perception, works constraints, insecurity and crime, differential perceptions by age, and the availability of organized activities in the parks. Residents from high-NSES reported fewer barriers to park use compared to residents from the other areas, who reported limitations such as less leisure time due to job constrains or perceived insecurity in parks. Senior participants reported that having parks with organized activities and a design oriented towards different age-groups are valuable. Our study shows consistency between the fewer and less intense use of parks registered in the middle and low-NSES neighborhoods, and the more barriers for PA reported in this areas during the qualitative analysis. Mixed-methods provided an insight of the potential causes leading to the differences in park use and PA within cities, which is essential in terms of environmental justice and health equity. Thus, a mixed-methods comprehensive approach to public health problems can help designing public policies addressing relevant factors related to urban health inequities. * Corresponding author. Campus Universitario - C/ 19, Av. de Madrid, Km 33,600, 28871, Alcalá de Henares, Madrid, Spain. E-mail address: pedro.gullon@uah.es (P. Gullón). https://doi.org/10.1016/j.healthplace.2021.102660 Received 30 March 2021; Received in revised form 23 July 2021; Accepted 19 August 2021 Available online 25 August 2021 1353-8292/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). M. Fontán-Vela et al. Health and Place 71 (2021) 102660 1. Introduction park use and population’s perceptions, an approach that considers the role of subjectivity is warranted. A contextual and comprehensive While cities have the potential to improve population’s well-being analysis is necessary to fully understand how structural and individual through a health-focused design, 23% of worldwide deaths that determinants are interrelated, and how they are influencing the use of occurred during 2012 were due to unhealthy environmental conditions, urban parks for PA. We thus ask: What is the relationship between such as a poorly built environment (Prüss-Ustün et al., 2016). An different socioeconomic contexts, and the use and perceptions of parks? extensive body of research acknowledges the role of neighborhood-level We draw on a mixed-methods approach, which, by combining quanti­ physical and social environments for health outcomes (Acevedo-Garcia tative and qualitative data, allows for a better understanding of complex et al., 2014; Diez Roux and Mair, 2010; Duncan and Kawachi, 2018; phenomena such as the relationship between social and urban dynamics Fullilove and Wallace, 2011; Hill et al., 2005; Sampson and Laub, 2003; and health outcomes (Cunningham-Myrie et al., 2019; Uijtdewilligen Stevenson et al., 2016). For instance, a closer residential proximity to et al., 2019); . green spaces has been associated with better cardiovascular and In this study, we aimed to analyze differences in the use of parks for immunity-related outcomes (Cavaleiro Rufo et al., 2020; Dalton and physical activity and the extent to which individuals from different so­ Jones, 2020), and self-perceived and mental health (Klompmaker et al., cioeconomic backgrounds perceived these spaces as an urban asset for 2019; Triguero-Mas et al., 2015). physical activity in the city of Madrid, Spain. Green spaces might improve health through different pathways, such as reducing harm (mitigating exposures to heat, noise, and air pollu­ 2. Methods tion), relieving mental and physiologic stress, and promoting healthy activities such as physical activity (PA) (Nieuwenhuijsen et al., 2017). 2.1. Study design However, some of the evidence related to green spaces is contra­ dictory, and not all types of them improve health through the same We designed a convergent-parallel mixed-methods study where we pathway. For example, the association between PA and green spaces collected and analyzed quantitative (systematic social observation in six depends on the setting or the study methodological design (Bancroft parks) and qualitative (residents’ perceptions) data separately, and then et al., 2015). Green spaces have been defined using different categories compared and contrasted the different results during the interpretation in urban research on greening and health. For instance, “green spaces” is (Bryman, 2006; Creswell and Plano Clark, 2011). Thereby, we gave usually used as a broad term that involves all types of green areas and equal weight to both data types. All the data collection, analyses, and elements that are available in the cities such as regular vegetation (e.g., interpretation processes were conducted between 2018 and 2019. We flower beds, trees), community gardens, or urban forests; meanwhile, conducted this study following the Declaration of Helsinki and received “urban parks” is a more specific concept that refers to those green areas ethics approval from the University of Alcalá (CEI/HU/2017/18). All with a minimum size (0.5 ha), vegetation, and walking paths around and participants provided written informed consent. through their area (WHO Regional Office for Europe, 2016), and that could count with other features such as sport courts, which could in­ 2.2. Study setting crease PA. In fact, different studies have shown a relationship between residential park proximity or density with an increase in PA (Akpinar, Our study area was the city of Madrid, Spain’s 3.2 million-resident 2016; Bancroft et al., 2015; Cohen et al., 2006; Durand et al., 2011; capital, divided into 21 districts and 128 neighborhoods and with one Papas et al., 2007; Young et al., 2014). of the highest increases in socioeconomic segregation in Europe (Leal Regarding the distribution of green spaces and parks, some neigh­ and Sorando, 2015). We selected three neighborhoods with different borhoods tend to be more deprived than others in terms of access to socioeconomic status to bring in an environmental equity lens. Briefly, green spaces and face acute environmental injustices (Anguelovski et al., we followed a two-step process: first, we divided, ranked, and 2020b; Langemeyer and Connolly, 2020). For instance, some areas with divide-classified into tertiles all neighborhoods by their low socioeconomic status have fewer and smaller parks, urban parks neighborhood-level socioeconomic status (NSES) (Rivera Navarro et al., with worse quality, and less accessible and safe in comparison with 2019) and selected nine neighborhoods (three per tertile). The NSES wealthier neighborhoods (Cohen et al., 2012; Engelberg et al., 2016; index was built using neighborhood-level variables related to available Jones et al., 2009; Schüle et al., 2019). data on the demographic and socioeconomic structure of the neigh­ However, the mere presence of green spaces is not always sufficient borhood such as the level of unemployment, people with a part-time or to create healthy environments. How populations interact with different temporary job, people with low educational level, people born in a features of the urban environment relies on a complex system where the foreign country, and households with a single parent and one or more economic and social structure of a certain area influence the social dy­ children. Second, we selected one neighborhood per tertile using namics that shape inhabitants’ behaviors and how they perceive these non-participant observation, leaving a final sample of one low-NSES features (Franco et al., 2015; Markevych et al., 2017; ). For example, (San Diego), one middle-NSES (El Pilar), and one high-NSES (Nueva several analyses have pointed out the discrepancy between objective España) neighborhoods. More details about the neighborhood selection measures of urban characteristics (e.g., accessibility or availability) and process can be found elsewhere (Rivera Navarro et al., 2019). how individuals perceive them (Brownson et al., 2009; Hinckson et al., At first, we sampled one park per neighborhood with similar size and 2017; Knobel et al., 2021; Kothencz and Blaschke, 2017; Ma et al., 2014; facilities (e.g., children’s playground or sports facilities). As the quali­ Orstad et al., 2017), and related those discrepancies to environmental tative strand was being conducted in parallel, three more parks (one per and health justice issues (Anguelovski et al., 2020a). Furthermore, social neighborhood) that complied with the previous criteria (similar size and groups with different sociodemographic characteristics, such as gender, facilities) were added to the sample as they were identified as an asset age, or ethnicity, experience the impact of exposure to green space for PA by the participants. differently, even when comparing them to more privileged groups living Parks selected had similar size (around 50000 square meters), in the same neighborhood (Kabisch et al., 2017; Pearce et al., 2018). counted with walking trails, and offered the same facilities like chil­ However, while some studies have started to offer a more contextualized dren’s playgrounds, wide-open spaces for informal games, or sports understanding of how the presence of urban parks combines with courts, and also had ornamental features like water fountains. Fig. 1 complex socioeconomic and political environments, there is still much shows the context of the city of Madrid and depicts pictures of some of space to further explore these dynamics. Our paper aims to fill this gap the parks included for the systematic social observation. with a novel focus on park perceptions, use, and physical activity. Thus, to understand the interaction between those factors related to 2 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 Fig. 1. Neighborhoods of the city of Madrid according to their socioeconomic status (SES), the final neighborhoods sample (A, B and C), and pictures of some of the parks selected for the systematic social observation within each neighborhood. 2.3. Quantitative strand: systematic social observation 2.3.3. Data analysis We exported quantitative data into two different datasets: number of 2.3.1. Audit tool visitors by age-group, gender, and number of people performing We obtained quantitative data on PA using the System for Observing different PA levels. Then, we performed a descriptive analysis of park Play and Recreation in Communities (SOPARC) (Mckenzie and Cohen, use by different age-group, gender, NSES, and the iSOPARC variables. 2006), a widely previously-used systematic social observation tool for We also analyzed differences in physical activity levels by NSES and assessing park and recreation areas (Ward et al., 2014). The Sport Fac­ gender. For these analyses, we used R software, version 4.0.3. ulty of the University of Oporto adapted the SOPARC methodology for an iPad application systematizing the data collection procedure (Santos et al., 2016). 2.4. Qualitative strand: perceptions of parks and physical activity 2.3.2. Registration procedure 2.4.1. Semi-structured interviews (SSIs) and focus groups (FGs) Within each park, we selected three target areas that represented To conduct the SSIs (n = 37) and FGs (n = 29), we identified and standard locations likely to provide opportunities to be physically selected participants based on the following criteria: sex, age, educa­ active, defined as areas with enough space that allow activities involving tional level, number of children, employment status, income, family different levels of PA (e.g., walking, running, or riding a bike). At least responsibilities related to children or grandchildren, years of residence one of these target areas had to have a children’s playground to capture in the neighborhood, and country of origin. A professional market potential patterns associated with childhood and with care. These target research agency recruited participants, who were contacted via tele­ areas were selected after the consensus of two researchers following the phone. The response rate was high; 92% of contacted residents enrolled same criteria defined by the SOPARC protocol. in the study. Participants received financial compensation of 25€ and a Each park was visited twice by a researcher: once on a working day bottle of extra virgin olive oil. Using a professional agency made it and once during the weekend. Each visiting day we set up a protocol impossible to include children users in our analysis. Participants lived in with the same four registration shifts, through which every 1-h shift was the three neighborhoods of the study area and were, at least, 40 years divided into 20 min periods to register PA levels within each target area. old. The age criterion was established for the larger urban health project In the end, each target area within each park was visited four times in which the qualitative strand of the study was embedded (Bilal et al., throughout the day, on two different days (eight times in total). The 2016). member of the team who carried out the systematic social observation In the SSIs, researchers interviewed (JRN, PCE, MGS, MSV, and MFT) registered the following variables about every person who crossed the 31 residents and 6 key informants. Key informants were school princi­ target area: pals and directors of healthcare centers who helped to explain the dy­ namics related to our study dimensions in each neighborhood. Table 1 • PA: low/medium/high. SOPARC protocol defines “high PA” as any shows the sociodemographic characteristics of participants. kind of activity that could increase the heart rate causing sweating. For the focus groups, we conducted 14 in the low-SES, 11 in the • Gender perceived by the observer: male or female. middle-SES, and 4 in the high-NSES. In total, 182 residents participated • Age-group perceived by the observer: children (0–12 years), ado­ in the 29 focus groups (Table 2). lescents (13–20 years), adults (21–59 years), or seniors (60 years and older). 2.4.2. Data collection • Moment of the day: morning, noon, afternoon, and evening. The SSIs and the FGs were conducted between 2016 and 2019. The • Day of the week: weekday or weekend. SSIs lasted approximately 60 min and were conducted by six re­ searchers. FGs took place between 2018 and 2019 and lasted approxi­ All parks included were visited on sunny days during May and June mately 90 min each. The number of participants within each FG ranged to avoid weather variations that could discourage residents from visiting from five to eight. We designed the SSIs and FGs guides (JRN, MFT, these spaces, such as rainy or extremely warm summer days. MGS, PCE, and MSV) including the following topics related to the neighborhood and parks: 1) general perception about their accessibility; 2) perception of their situation and condition); 3) perception of facilities 3 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 Table 1 conducted in the high SES neighborhood. The number of FGs conducted Sociodemographic characteristics of semi-structured interviews participants (n in the low-SES neighborhood was higher due to the higher proportion = 31) and key informants (n = 6) participating in the semi-structured interviews (30%) of people born in a foreign country in this area, to better show according to neighborhood socioeconmic status. their different and diverse perceptions as sub-groups. Neighborhood The SSIs and FGs topic guides are included in the supplementary Low SES (n = Medium SES (n = High SES (n = material (S1). 12) 12) 13) Gender (n, %) 2.4.3. Data analysis Female 5 (41.7) 7 (58.3) 6 (46.2) All SSI and FG were professionally transcribed and then checked by Male 7 (58.3) 5 (41.7) 7 (53.8) four researchers (JRN, MGS, PCE, and MSV). Direct quotes were trans­ Age (n, %) lated from Spanish to English following a widely-used three-step pro­ 40-49 4 (33.3) 5 (41.7) 5 (38.4) cedure (Biering-Sørensen et al., 2011). We used the principles of 50-59 2 (16.7) 2 (16.7) 3 (23.1) 60-69 1 (8.3) 4 (33.3) 4 (30.8) constructivist grounded theory (Charmaz, 2006; Glaser, 2003) and ≥70 5 (41.7) 1 (8.3) 1 (7.7) included the complete dataset in this analysis (e.g., entire transcripts of Educational level (n, %) all 37 SSIs and 29 FGs). The research team coded the transcribed text ≤ Primary school or less 9 (75.0) 2 (16.7) – line-by-line using open coding. Constant comparison (Glaser, 2003) of Secondary school 1 (8.3) 6 (50.0) 2 (15.4) Tertiariy achool 2 (16.7) 4 (33.3) 11 (84.6) emerging codes allowed us to consolidate several categories and Employment status (n, %) sub-categories, which sometimes, although not always, matched the Working (full-time or 5 (41.7) 7 (58.4) 9 (69.2) core topics about the research question previously determined by the part-time) researchers. Then, categories were refined into broader themes to detect Unemployed 1 (8.3) – – conceptual similarities, refine differences between categories, and Retired 6 (50.0) 4 (33.3) 4 (30.8) Housewives – 1 (8.3) – discover discourse patterns. We used ATLAS.ti-8 software to manage the Country of origin (n, %) analytical process by four team members with different backgrounds Spain 8 (66.7) 12 (100.0) 13 (100.0) (social sciences and public health). These four team members compared Other 4 (33.3) – – codes as well as the resultant categories, ensuring the criteria of credi­ Living arrangement (n, %) Living alone 1 (8.3) 4 (33.3) 2 (15.4) bility (reliability/validity) (Dahlgren et al., 2004). Differences between Cohabiting 11 (91.7) 8 (66.7) 11 (84.6) resultant categories were resolved by a discussion and further clarifi­ SES: socioeconomic status cation among the team members. 2.5. Mixed-methods analysis Table 2 Sociodemographic characteristics of focus group participants (n = 182) ac­ We followed a convergent-parallel approach to collect and analyze cording to neighborhood socioeconomic status. both the quantitative and the qualitative data (Bryman, 2006; Creswell Neighborhood and Plano Clark, 2011; Edmonds and Kennedy, 2017). Both quantitative and qualitative strands of the study were conducted in parallel, therefore Low SES (n = Medium SES (n = High SES (n = two separate analysis were performed: the quantitative one after the 90) 65) 27) collection of data on park use, and the qualitative one after conducting Gender (n, %) the SSI and FG, and codifying the perceptions and discourses of the Female 52 (57.8) 44 (35.4) 15 (55.6) Male 38 (42.2) 21 (64.6) 12 (44.4) participants. As the data of both strands were available, the analysis Age (n, %) were conducted separately, and the results are presented separately. 40-49 23 (25.5) 18 (27.7) 2 (7.4) Afterwards, we combined the evidence of each strand of the study 50-59 36 (40.0) 24 (36.9) 11 (40.7) (quantitative and qualitative) to compare them and detect areas of 60-69 25 (27.7) 12 (18.5) 5 (18.5) ≥70 6 (6.8) 11 (16.9) 9 (33.4) convergence or divergence in the findings. Following this approach, we Educational level (n, %) did not give more relevance to either strand of the study, but we focused ≤ Primary school or less 20 (22.2) 7 (10.7) 2 (7.4) on whether there was consistency between both quantitative and qual­ Secondary school 33 (36.7) 27 (41.6) 2 (7.4) itative data, and we tried to understand the consistency or discrepancy Tertiariy achool 37 (41.1) 31 (47.7) 23 (85.2) based on the information gathered in each strand of the study, also by Employment status (n, %) Working (full-time or 60 (66.7) 39 (60.0) 18 (66.7) comparing them with the available evidence in the research literature. part-time) We organized the results section as follows: first, we summarize the Unemployed 9 (10.0) 7 (10.7) – results from the quantitative analysis drawn from our systematic social Retired 17 (18.9) 18 (27.7) 9 (29.7) observation; and, second, we present the qualitative results by the most Housewives 4 (4.4) 1 (1.6) 1 (3.6) relevant topics related to park use and physical activity detected during Country of origin (n, %) Spain 62 (68.9) 49 (75.4) 27 (100.0) the analysis. In the discussion section, we explain the convergence and Other 28 (31.1) 16 (24.6) – divergence areas of both strands and the potential explanations of the Living arrangement (n, %) observed phenomena. Living alone 25 (27.7) 12 (18.5) 5 (18.5) Cohabiting 65 (72.3) 53 (81.5) 22 (81.5) 3. Results and resources related to parks and physical activity that take place in In this section we first present park use trends building on our each neighborhood. We completed SSIs and FGs after reaching satura­ quantitative analyses, and then report on the five most salient factors, tion point when perceptions shared started to be repetitive, and no new for park use, and especially for physical activity, according to our core themes and issues in participants’ answers or discussions emerged qualitative analysis: a) Park maintenance and area perception; b) Work (Axelsson et al., 2015). The saturation point was reached earlier in the constraints; c) Insecurity and crime; d) Differential perceptions by age; SSIs and FGs conducted in the high-SES neighborhood because of their e) Availability of organized activities in the parks. more homogeneous resident profile (i.e., mostly high-skilled workers and high-income residents), which explains the fewer number of FGs 4 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 3.1. Quantitative results: park use and physical activity by neighborhood Table 4 socioeconomic status and sociodemographic characteristics Differences on physical activity performance according to the interaction be­ tween neighborhood socioeconomic status and gender (N = 10810). We registered 10,810 individual observations throughout the quan­ Low PA (n, %) Medium PA (n, %) High PA (n, %) Total (n, %) titative data collection period. Table 3 presents the quantitative data High-NSES related to park use and PA according to NSES, gender, day of the week, Female 779 (30.0) 1588 (61.1) 232 (8.9) 2599 (52.4) and moment of the day. Male 600 (25.5) 1400 (59.4) 357 (15.1) 2357 (47.6) In general terms, the number of observations was similar in a regular Medium-NSES Female 326 (23.8) 935 (68.2) 110 (8.0) 1371 (47.9) workday and weekend, although some differences were registered. More Male 314 (21.1) 1019 (68.3) 158 (10.6) 1491 (52.1) people performed medium PA in working days as compared to those Low-NSES during the weekend (68.2% vs 62.8%), but more people performing high Female 332 (24.3) 996 (72.9) 39 (2.8) 1367 (45.7) PA were registered during the weekend (6.0% vs 11.7%). According to Male 421 (25.9) 1146 (70.5) 58 (3.6) 1625 (54.3) the moment of the day, more people visited these spaces in the afternoon NSES: neighborhood socioeconomic status; PA: physical activity. and the evening (32.7% and 39.3%, respectively). According to the NSES, we registered more visits in parks located in children was found within the medium-NSES (14.4%). Adolescents were the high-NSES areas (4956) compared to low-NSES areas (2992); in fact, the group with the lowest proportion registered across the three NSES, 45.8% of all visits happened in the high-NSES parks. Regarding levels of but this was more remarkable in the high-NSES (4.8%). The proportion PA, most of the people performed medium PA (7084, 65.5%). Some of seniors was higher in the medium-NSES area (31.2%). differences in PA according to the NSES were observed. For example, the The proportion of males and females was similar in the children and proportion of visitors performing high PA followed a social gradient: adolescents groups, but some differences were registered in the two That proportion was higher in the high-NSES (11.9%) compared to the other groups: more proportion of females compared to males were found medium and low-NSES (9.3% and 3.2%). in adults (54.9% vs 50.4%); while the opposite was observed in the se­ According to gender, the proportion of male and females visitants nior group (23.8% vs 27.3%). was similar (50.6% and 49.4%, respectively). More females performed Across all groups except for the children, a higher use of parks was low PA compared to men (26.9% vs 24.4%), meanwhile, a higher pro­ registered in the working days. However, there were higher differences portion of men performed high PA (7.1% vs 10.5%). Table 4 shows in the observed proportion of children between a working day and a differences in PA performed according to both the NSES and gender. weekend (9.8% vs 17.7%) compared to the other age-groups. Differences in the proportion by gender performing high PA differed by NSES: these were higher in the high-NSES (15.1% males vs 8.9% 3.2. Qualitative results: reported factors for park use and physical activity females) compared to the low-NSES where proportions were more comparable (3.6% males vs 2.8% females). Fig. 2 shows how differences During the SSIs and FGs, participants spoke about their neighbor­ in PA performed by NSES varied by gender: within males, differences by hoods, specifically about the facilities located within them, such as NSES were wider for high PA, while in females the relation was the parks, and related to physical activity (e.g., sports, walking, etc.). The opposite, differences were wider for the low PA. following issues were reported as influencing park use for physical Table 5 shows the distribution by age group. Adults were the group activity: that more frequently visited parks (52.6%). The highest proportion of 3.2.1. Park maintenance and NSES area perception People from the high-NSES area argued that both the facilities Table 3 (including parks) and clean air were ideal for walking and sports. They Differences in physical activity level performance by park visitors according to neighborhood socioeconomic status, gender, day of the week, and the moment mentioned also how parks were integrated with the rest of the built of the day (N = 10810). environment, as a broader perception of the public space, was important to encourage park use for walking or PA: Low PA (n, Medium PA (n, High PA (n, Total (n, %) %) %) %) The neighborhood attracts [users] because there are green spaces. Yes, the Neighborhood SES neighborhood attracts, yes, yes. I think the air is a little cleaner than El High-NSES 1379 (27.8) 2988 (60.3) 589 (11.9) 4956 Retiro (one of the biggest parks of the city), because El Retiro is in the (45.8) center of Madrid (FG. Male and female, 45–55 years old. High income. Medium- 640 (22.4) 1954 (68.3) 268 (9.3) 2862 NSES (26.5) High-NSES) Low-NSES 753 (25.2) 2142 (71.6) 97 (3.2) 2992 (27.7) Specifically, I would tell you this whole neighborhood and this whole area, Gender including its parks, is ideal to walk around and do any type of physical Female 1437 (27.0) 3519 (65.9) 381 (7.1) 5337 activity, go out to walk. You can walk and do any type of physical activity (49.4) in neighborhood parks and nearby parks in other neighborhoods (SSI. Male 1335 (24.4) 3565 (65.1) 573 (10.5) 5473 Active male. 53 years old. Without children. High income. High- (50.6) Day of the week NSES). Working day 1400 (25.8) 3707 (68.2) 326 (6.0) 5433 (50.3) However, people from the middle and low-NSES neighborhoods re­ Weekend 1372 (25.5) 3377 (62.8) 628 (11.7) 5377 ported that their parks were very dirty because of incivilities and also (49.7) because they felt abandoned by the public administration: Moment of the day Morning 63 (6.5) 816 (84.6) 86 (8.9) 965 (8.9) … The same thing happens in the parks. There are very dirty parks, very Noon 916 (25.9) 2363 (66.9) 255 (7.2) 3534 dirty … most of them, 99% … and no … rubbish bins are broken and they (32.7) aren’t replaced …. and they aren’t emptied. But I think the problem is not Afternoon 556 (27.0) 1395 (67.7) 109 (5.3) 2060 (19.1) so much that they do not do it, but rather that there is little civic re­ Evening 1237 (29.1) 2510 (59.0) 504 (11.9) 4251 sponsibility (FG. Male and female, 45–55 years old. Stable workers. (39.3) Medium-NSES) SES: socioeconomic status; PA: physical activity. 5 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 Fig. 2. Differences in the number of observations by PA and gender within each neighborhood according to its socioeconomic status (SES) (N = 10810). Unemployed and precarious workers. Some immigrant residents. Table 5 Low-NSES). Differences on visits to parks of different age-groups according to neighborhood SES, gender, day of the week or the moment of the day (N = 10810). Children Adolescents Adults Senior Total 3.2.3. Insecurity and crime (n, %) (n, %) (n, %) (n, %) (n, %) People from the low-NSES neighborhood reported feelings of inse­ Neighborhood SES curity within the neighborhood as an important barrier for park use. High SES 692 (14.0) 237 (4.8) 2865 1162 4956 This was associated by the participants with certain social groups who (57.8) (23.4) (45.8) used parks inappropriately. Specifically, three factors were highlighted Medium 411 (14.4) 312 (10.9) 1245 894 2862 by respondents: 1) loose dogs; 2) people drinking on the streets (young SES (43.5) (31.2) (26.5) Low SES 380 (12.7) 326 (10.9) 1576 710 2992 people and drug users in particular), and 3) young LatinX residents (52.7) (23.7) (27.7) (mainly gangs linked to episodes of violence). Gender Although there were public parks in the low-SES neighborhood, the Female 734 (13.7) 404 (7.6) 2928 1271 5337 above-mentioned factors caused many neighbors to perceive that they (54.9) (23.8) (49.4) Male 749 (13.7) 471 (8.6) 2758 1495 5473 could not use them and to use other parks located in other (50.4) (27.3) (50.6) neighborhoods: Day of the week Working 531 (9.8) 509 (9.3) 2932 1461 5433 There is a lot of crime, there are a lot of bad people in the neighborhood day (54.0) (26.9) (50.3) who weren’t there before. In other words, you can’t walk in the park after Weekend 952 (17.7) 366 (6.8) 2754 1305 5377 certain hours. And you always have to be accompanied. (FG. Seniors (51.2) (24.3) (49.7) older than 65 years old. Retired. Low-NSES). Moment of the day Morning 34 (3.5) 135 (14.0) 580 216 965 “… But usually we have to get the dogs into the car and take them to (60.1) (22.4) (8.9) Noon 441 (12.5) 198 (5.6) 1691 1204 3534 another park … in our neighborhood, it’s impossible because there is a (47.8) (34.1) (32.7) group of kids who are 20 years old and who usually bring loose pit bulls” Afternoon 210 (10.2) 216 (10.5) 1265 369 2060 (FG. Females from 45 to 55 years old, with different family situations (61.4) (17.9) (19.1) -with or without children, married, divorced, single-, some immi­ Evening 798 (18.8) 326 (7.7) 2150 977 4251 grants; low-NSES neighborhood) (50.6) (22.9) (39.3) SES: Socioeconomic status; PA: physical activity levels. 3.2.4. Differential perceptions by age Our parks are not cleaned. We are also to blame, there is no civic re­ In the medium-NSES neighborhood, older adults reported that peo­ sponsibility. People throw bags on the ground instead of in the bins. The ple across all age-groups used parks for walking, but this was more parks are not cleaned. (SSI. Widow female. 83 years old. She lives characteristic of the senior group: alone. Low-NSES). People go out to walk. A lot of people go for a walk in the park, at any time of the day. Cold or hot, rain or snow, they are there walking every day. 3.2.2. Work constraints There are people of all ages, but mostly seniors (FG. Housewives, older Residents reported as a relevant factor the nature of the job and the than 65 years old. Medium-NSES). availability of leisure-time. People from the low-NSES highlighted One relevant reason to use these areas reported by seniors from the feeling physically exhausted after work, which discouraged them from medium-NSES neighborhood was its role as a place where their grand­ using parks for PA. Additionally, they reported that long-lasted shifts left children could socially interact with their friends: them without enough leisure-time to visit parks: Since I have grandchildren, I usually go down to the park. You have to go An example, you wake up at 6 a.m. to go to work and you get to work and down to a park, you no longer go for a walk, you go to the park where you leave at 8 p.m. and you say am I going to a park to do exercise? there are children who interact with children (FG. Housewives, older Because tomorrow you wake up at 6 a.m. again … Then, when you have than 65 years old. Medium-NSES). Sunday off, you say I’m going to rest (FG. Male, 45–55 years old. 6 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 3.2.5. Availability of organized activities in the parks During the interviews and focus groups, residents from the low-NSES The variety of organized activities to perform PA in groups or the reported the presence of incivilities and insecurity as barriers to visit social cohesion within a neighborhood that leads to the spontaneous these spaces. These barriers have been already reported in previous organization of these activities were valued as important characteristics studies (Lapham et al., 2016; Leslie et al., 2010; McCormack et al., that make parks more attractive for visits and PA, especially in the 2010). Moreover, these barriers have been shown to have a particularly medium-NSES neighborhood: strong impact on female and working-class residents’ use of parks (Cronan et al., 2008; Wilbur et al., 2002; Wilson et al., 2004). Under­ People do the cholesterol route, which is a path around the neighborhood standing the roots of insecurity and crime is thus important to changing that also passes through several parks. There are a lot of seniors, women neighbors’ perceptions about safety. One potential explanation is that walking around the neighborhood everywhere (SSI. Key informant. perceptions about neighborhood insecurity are also driven by the social Health Center. Medium-NSES). construction of stereotypes about racial and ethnic minorities, specially There is an esplanade next to the pond where we do calisthenics. I already within those areas with a higher percentage of these social groups said that I have been doing it for 8 years and there are people who have (Quillian and Pager, 2001). These stereotypes might have more influ­ been there for longer and we do calisthenics there. What has happened is ence in these neighborhoods, which could be related to the less park use that our group has acquired a life of its own (FG. Housewives, older observed in low-NSES and the perceptions reported by their residents. than 65 years old. Medium-NSES). Other studies describe that gentrification has an impact on social sta­ bility and cohesion, which could increase conflict and sense of being As one participant from the medium-NSES reported, these types of excluded from space (Oscilowicz et al., 2020; Van Welsem et al., 2006). activities were so important that the neighbors opposed to their modi­ Another aspect that limited the use of the parks in Madrid as reported by fication or cancellation: residents was dirtiness and litter, especially in the low and middle-NSES, in line with other studies on the role of poor aesthetic or maintenance on A support staff from the City Council wanted us to modify the schedule and I told him. “We already have a life of our own and we do calisthenics urban park use (McCormack et al., 2010), and on the lower public re­ sources dedicated to park clean up or upgrading in lower-income at 9:30 and 10:30 and then if there are volunteers who want to do another schedule, they can do it, but we are here" (FG. Retired male over neighborhoods (Rigolon, 2016). Residents of the low-NSES reported how their job characteristics and 65 years old. Medium-NSES). schedule constraints affect their leisure-time availability or tiredness, leading to less motivation or ability for physical activity. Other studies 4. Discussion have pointed out the relevant role of occupational determinants on sedentarism (Beenackers et al., 2012; Seiluri et al., 2011). Upgrading or Our mixed-methods study showed consistency between the quanti­ increasing the availability of green spaces could not be a sufficient factor tative (systematic social observation) and the qualitative strand (resi­ to increase physical activity, and requires more comprehensive policies dents’ perceptions). Parks within the high-NSES neighborhood were to improve socioeconomic opportunities and job conditions among the more frequently used by a higher proportion of visitants performing most disadvantaged groups. Furthermore, a more intense PA were high PA levels, and residents from this high-NSES perceived fewer bar­ registered in weekend days, so policies aim to increase population riers to use parks. We found more males perform high PA compared to leisure-time could lead both to increase regular PA, benefiting the more females, and differences were more pronounced in the high-NSES. Fewer deprived. seniors and children visited parks in the low-NSES compared to the other neighborhoods, where seniors and female participants reported an un­ 4.2. Differences in park use by gender in the different NSES safe park environment. Adolescents were the group with fewer visits to parks across all NSES. Generally, female residents visit less these spaces and perform lower Extensive research shows that physical activity relates to better levels of physical activity compared to men (Cunningham-Myrie et al., health outcomes and that this relationship appears to be not dose- 2019; Derose et al., 2017; Knapp et al., 2019; Vaughan et al., 2018). Our dependent: any physical activity provides some health benefit (War­ results are consistent with this existing evidence: more females were burton and Bredin, 2017). Nevertheless, physical activity levels are not observed performing low physical activity and performing less high equally distributed across different social groups (Aleksovska et al., physical activity compared to males. Female residents suffer more often 2019; O’Donoghue et al., 2018; Willey et al., 2010). Considering these from different types of violence in the streets, so they experience more inequalities is key in public health research and the design of environ­ frequently insecurity, as the analysis of the perceptions showed in low ments that mitigate health inequities (Anguelovski et al., 2020a,b). and middle-NSES, and are thus more discouraged from using green Our study provides data that could explain differences in park use public spaces (Oscilowicz et al., 2020). Another explanation could be and physical activity across different socioeconomic contexts. The rea­ that female and male residents use parks for different purposes due to sons reported by the participants allow for developing an integrated gender roles: men usually attend these spaces to perform higher physical framework for better understanding the relationship between parks, use, activity levels, while women are caring for their children or grand­ and physical activity at different levels: park-level (aesthetics or orga­ children (Cohen et al., 2019; Derose et al., 2017). Furthermore, our nized activities), individual-level (gender roles or job characteristics), analysis showed that differences between males and females, and within and contextual-level (neighborhood perceptions of safety concerns). both groups, were influenced by the NSES, especially in the female Here, we discuss the main issues reported in the qualitative strand, and group; more females were performing low physical activity in the weather their convergence or divergence with the quantitative data: high-NSES neighborhoods. Some of these females were caring for their children during the systematic observation, so the more presence of 4.1. Less park use in low-NSES and barriers reported by their inhabitants them performing low physical activity in privileged areas could be related to having less economic constraints, and thus more possibilities We found that parks located in high-NSES were more used compared to take parental leave, compared to females in more underprivileged to those located in more disadvantaged areas, similar to previous studies areas with more unstable economic conditions. (Cohen et al., 2016; Leslie et al., 2010). Previous studies also found a negative association between neighborhood deprivation and green 4.3. Park use and perceptions by age group space accessibility, quality, or other characteristics (Cohen et al., 2012; Hoffimann et al., 2017; Vaughan et al., 2018; Wen et al., 2013). How individuals interact with green spaces, and their perceptions 7 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 about them, change during their life course (O’Donoghue et al., 2018). registered could be explained by the other elements analyzed Participants in our study reported that older adults organized activities throughout the manuscript. The systematic observation was performed within the medium-NSES parks to perform physical activity. Other following a specific schedule which could not be the most suitable for studies have found that these organized activities are correlated with some age-group according to how they use parks (e.g., adolescents). higher use of these urban assets (Cohen et al., 2012), and might ingrease Another potential limitation is the availability of data on ethnic origin in social cohesion (Jennings and Bamkole, 2019). A lower proportion of the quantitative information. However, as the SOPARC relies on obser­ older adults was observed in the low-NSES parks, which could be vation, a decision to include ethnicity could have led to a misclassifi­ associated with the insecurity as well as the lower quality and mainte­ cation of park users. Last, the sample size of parks does not allow us to nance that they reported during the qualitative data collection. make inferences about park use in the entire city of Madrid. Neverthe­ Adolescents were the age-group that was less likely to visit parks. less, several criteria to select parks and participants were followed to Other studies have shown similar results (Babey et al., 2015). Some of have a variety of green spaces which enable us to catch potential dif­ the facilities that adolescents value such as skate parks or courts were ferences both on park use and perceptions about parks and physical present only in the high-NSES parks, as has been reported in other activity. studies on the quality differences of green spaces within the cities (Edwards et al., 2015; Floyd et al., 2011). Besides the facilities, another 5. Conclusions important barrier described in the literature, and reported during the qualitative research, is park safety (Akpınar, 2020; Babey et al., 2015), Our study showed a differential pattern on park use and physical although our qualitative groups did not include adolescents. activity, and analyzed the different residents’ perceptions about these spaces and physical activity according to neighborhood socioeconomic 4.4. Strengths and limitations status (NSES). Parks were more visited in the high-NSES neighborhood, where a higher proportion of people performed high physical activity (as The main strength of the study is the use of a mixed-methods compared with residents in the middle- and low-NSES areas). In the approach, which allowed us to have a more comprehensive under­ latter, residents reported several barriers to park use, such as poor standing of park use and their role in promoting physical activity in the maintenance and unsafety. As compared to male residents, a lower city of Madrid, and how individual and neighborhood characteristics proportion of female residents performed high physical activity, who have an influence on their use and the perceptions about these spaces. also reported insecurity concerns. The convergent-parallel mixed- This methodology combines different sources of information that pro­ methods analysis provided valuable insights on the potential causes that vide potential explanations to the observed phenomena and deepens could explain the quantitative evidence, pointing out potential path­ into the links between park, use, physical activity, and social behavior ways to understand the further implications of inequalities in park use in within cities. The evidence generated through this process could be terms of environmental and health inequities. useful to follow a more population approach framework as Geoffrey A better understanding of how individual factors (e.g, gender or age) Rose proposed (Rose, 1985), focusing on those factors that condition and contextual ones (e.g., neighborhood safety or socioeconomic status) park use and physical activity at the population level. In this regard, influence the unequal relationship between parks, their use, and phys­ studies on the relationship between neighborhoods and health have ical activity, and how those shape individual perceptions, is key to pointed out the importance of contextual variables on the individual building healthier neighborhood environments for all. These results call level (Franco et al., 2015). Although the study was conducted in the city for public action and investment in public green spaces to increase their of Madrid, the variety of participants and the richness of their discourses maintenance and availability of organized activities, improving their and perceptions provided information about the potential pathways and attractiveness for all ages, and to address residents’ perceptions related barriers for active park use that could be generalized to other cities with to security and those contextual factors that constrain their use, espe­ similar urban dynamics. Thus, collecting both individual and contextual cially in underserved urban areas. variables could lead to a better understanding of urban health and a better design of public policies aimed to increase park use and physical Declaration of competing interest activity within cities. Madrid is also a large, dense, global city where public and green space use is no different from any other place with The authors declare no conflict of interest. similar characteristics. Visiting each park twice and across different moments of the day Acknowledgments helped to reduce the confusion related to the variability on park use according to the moment of the week or day. Another strength was our Researchers would like to thank all the participants of the interviews ability to include parks and individuals from neighborhoods of different and focus groups and all the interviewers. We appreciate the contribu­ socioeconomic status to detect the potential effect of contextual tion of all the qualitative research team of the urban health project that variables. contribute to collect the data related to physical activity and parks (PCE, The study also presents several limitations. Not all park users might MGS, MSV, and MFT). Also, we would like to thank the team from the live in the park area; however, the convergence with the discourses of Sport Faculty of the University of Oporto that developed the iSOPARC neighbors in the qualitative analysis reduces the risk of this potential application. We would also thank all the nameless pedestrians who were bias. Data obtained from the systematic social observation tool was observed during the systematic observation for their unselfish aggregated and not analyzed at the individual level, which would allow contribution. a more accurate analysis of park active use. However, our aggregated data showed differences observed in other studies. Another limitation Appendix A. Supplementary data was the exclusion of some age groups from the qualitative interviews (e. g., kids or adolescents). Thus, further studies are needed to understand Supplementary data to this article can be found online at https://doi. which park characteristics are valued by these age-groups. Potential org/10.1016/j.healthplace.2021.102660. differences in the quality of parks could lead to differences in the use pattern. However, this was minimized by selecting parks with similar Funding characteristics (e.g.: size or facilities provision) in each neighborhood. Furthermore, we considered that by selecting parks with similar char­ This project was funded by the Instituto de Salud Carlos III, Sub­ acteristics in terms of size and facilities, the differences in park use dirección General de Evaluación y Fomento de la Investigación, 8 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 Government of Spain (PI18/00782) and by the Fondo Europeo de Charmaz, K., 2006. Constructing Grounded Theory: a Practical Guide through Qualitative Analysis. London. Desarrollo Regional (FEDER), by the European Research Council under Cohen, D., Sehgal, A., Williamson, S., Sturm, R., McKenzie, T., Lara, R., Lurie, N., 2006. the European Union’s Seventh Framework Programme (FP7/ Park Use and Physical Activity in a Sample of Public Parks in the City of Los Angeles. 2007–2013/ERC Starting Grant Heart Healthy Hoods Agreement no. RAND Corporation. https://doi.org/10.7249/TR357. 623 336893) and by the Spanish Ministry of Economy and Competi­ Cohen, D.A., Han, B., Derose, K.P., Williamson, S., Marsh, T., Rudick, J., McKenzie, T.L., 2012. Neighborhood poverty, park use, and park-based physical activity in a tiveness, Government of Spain (CSO2016-77257-P). Southern California city. Soc. Sci. Med. 75, 2317–2325. https://doi.org/10.1016/j. socscimed.2012.08.036. Cohen, D.A., Han, B., Nagel, C., Harnik, P., Mckenzie, T.L., Evenson, K.R., Marsh, T., Authors contribution Williamson, S., Vaughan, C., Katta, S., 2016. The first national study of neighborhood parks: implications for physical activity. Am. J. Prev. Med. 51, M.F-V, J.R-N, P.G. and M.F. conceptualized and designed the study; 419–426. https://doi.org/10.1016/j.amepre.2016.03.021. Cohen, D.A., Han, B., Park, S., Williamson, S., Derose, K.P., 2019. Park use and park- M.F-V and J.R-N prepared the data and wrote the first draft; P.G., M.F., I. based physical activity in low-income neighborhoods. J Aging Phys 27, 334–342. A. and J.D., critically revised the paper for intellectual content; J.R-N, P. https://doi.org/10.1123/japa.2018-0032. G. and M.F. performed project supervision and administration, and Creswell, J., Plano Clark, V.L., 2011. Designing and Conducting Mixed Methods managed funding acquisition. Research. Los Angeles, California. Cronan, M.K., Shinew, K.J., Schneider, I., Wilhelm Stanis, S.A., Chavez, D., 2008. Physical activity patterns and preferences among latinos in different types of public References parks. J. Phys. Activ. Health 5, 894–908. https://doi.org/10.1123/jpah.5.6.894. Cunningham-Myrie, C.A., Royal-Thomas, T.Y.N., Bailey, A.E., Gustat, J., Theall, K.P., Harrison, J.E., Reid, M.E., 2019. Use of a public park for physical activity in the Acevedo-Garcia, D., McArdle, N., Hardy, E.F., Crisan, U.I., Romano, B., Norris, D., Caribbean: evidence from a mixed methods study in Jamaica. BMC Publ. Health 19, Baek, M., Reece, J., 2014. The child opportunity index: improving collaboration 1–12. https://doi.org/10.1186/s12889-019-7247-6. between community development and public health. Health Aff. 33 https://doi.org/ Dahlgren, L., Emmelin, M., Winkvist, A., 2004. Qualitative Methodology for 10.1377/hlthaff.2014.0679, 1948–1957. International Public Health. Umea Universitet. Akpinar, A., 2016. How is quality of urban green spaces associated with physical activity Dalton, A.M., Jones, A.P., 2020. Residential neighbourhood greenspace is associated and health? Urban For. Urban Green. 16, 76–83. https://doi.org/10.1016/j. with reduced risk of cardiovascular disease: a prospective cohort study. PloS One 15, ufug.2016.01.011. e0226524. https://doi.org/10.1371/journal.pone.0226524. Akpınar, A., 2020. Investigating the barriers preventing adolescents from physical Derose, K.P., Han, B., Williamson, S., Cohen, D.A., 2017. Gender disparities in park use activities in urban green spaces. Urban For. Urban Green. 53, 126724 https://doi. and physical activity among residents of high-poverty neighborhoods in Los Angeles. org/10.1016/j.ufug.2020.126724. Physiol. Behav. 176, 139–148. https://doi.org/10.1016/j.physbeh.2017.03.040. Aleksovska, K., Puggina, A., Giraldi, L., Buck, C., Burns, C., Cardon, G., Carlin, A., Diez Roux, A.V., Mair, C., 2010. Neighborhoods and health. Ann. N. Y. Acad. Sci. 1186, Chantal, S., Ciarapica, D., Colotto, M., Condello, G., Coppinger, T., Cortis, C., 125–145. https://doi.org/10.1111/j.1749-6632.2009.05333.x. D’haese, S., Craemer, M., De Blasio, A., Di Hansen, S., Iacoviello, L., Issartel, J., Duncan, D.T., Kawachi, I., 2018. Neighborhoods and Health. Oxford University Press. Izzicupo, P., Jaeschke, L., Kanning, M., Kennedy, A., Ling, F., Luzak, A., https://doi.org/10.1093/oso/9780190843496.001.0001. Napolitano, G., Nazare, J.A., Perchoux, C., Pischon, T., Polito, A., Sannella, A., Durand, C.P., Andalib, M., Dunton, G.F., Wolch, J., Pentz, M.A., 2011. A systematic Schulz, H., Sohun, R., Steinbrecher, A., Schlicht, W., Ricciardi, W., Macdonncha, C., review of built environment factors related to physical activity and obesity risk: Capranica, L., Boccia, S., 2019. Biological determinants of physical activity across implications for smart growth urban planning. Obes. Rev. 12, e173–e182. https:// the life course: a “determinants of diet and physical activity” (dedipac) umbrella doi.org/10.1111/j.1467-789X.2010.00826.x. systematic literature review. Sport. Med. - Open 5, 1–26. https://doi.org/10.1186/ Edmonds, A.W., Kennedy, T.D., 2017. Convergent-parallel approach. In: An Applied s40798-018-0173-9. Guide to Research Designs: Quantitative, Qualitative, and Mixed Methods. SAGE Anguelovski, I., Brand, A.L., Connolly, J.J.T., Corbera, E., Kotsila, P., Steil, J., Garcia- Publications, Inc, 2455 Teller Road, Thousand Oaks California 91320, pp. 181–188. Lamarca, M., Triguero-Mas, M., Cole, H., Baró, F., Langemeyer, J., del Pulgar, C.P., https://doi.org/10.4135/9781071802779.n15. Shokry, G., Sekulova, F., Argüelles Ramos, L., 2020a. Expanding the Boundaries of Edwards, N., Hooper, P., Knuiman, M., Foster, S., Giles-Corti, B., 2015. Associations Justice in Urban Greening Scholarship: Toward an Emancipatory, Antisubordination, between park features and adolescent park use for physical activity. Int. J. Behav. Intersectional, and Relational Approach. Ann. Am. Assoc. Geogr. 110, 1743–1769. Nutr. Phys. Activ. 12, 1–10. https://doi.org/10.1186/s12966-015-0178-4. https://doi.org/10.1080/24694452.2020.1740579. Engelberg, J.K., Conway, T.L., Geremia, C., Cain, K.L., Saelens, B.E., Glanz, K., Frank, L. Anguelovski, I., Triguero-Mas, M., Connolly, J.J., Kotsila, P., Shokry, G., Pérez Del D., Sallis, J.F., 2016. Socioeconomic and race/ethnic disparities in observed park Pulgar, C., Garcia-Lamarca, M., Argüelles, L., Mangione, J., Dietz, K., Cole, H., quality. BMC Publ. Health 16, 1–11. https://doi.org/10.1186/s12889-016-3055-4. 2020b. Gentrification and health in two global cities: a call to identify impacts for Floyd, M.F., Bocarro, J.N., Smith, W.R., Baran, P.K., Moore, R.C., Cosco, N.G., socially-vulnerable residents. Cities Heal. 4, 40–49. https://doi.org/10.1080/ Edwards, M.B., Suau, L.J., Fang, K., 2011. Park-based physical activity among 23748834.2019.1636507. children and adolescents. Am. J. Prev. Med. 41, 258–265. https://doi.org/10.1016/j. Axelsson, L., Klang, B., Lundh Hagelin, C., Jacobson, S.H., Gleissman, S.A., 2015. amepre.2011.04.013. Meanings of being a close relative of a family member treated with haemodialysis Franco, M., Bilal, U., Diez-Roux, A.V., 2015. Preventing non-communicable diseases approaching end of life. J. Clin. Nurs. 24, 447–456. https://doi.org/10.1111/ through structural changes in urban environments. J. Epidemiol. Community Health jocn.12622. 69, 509–511. https://doi.org/10.1136/jech-2014-203865. Babey, S.H., Tan, D., Wolstein, J., Diamant, A.L., 2015. Neighborhood, family and Fullilove, M.T., Wallace, R., 2011. Serial forced displacement in American cities, individual characteristics related to adolescent park-based physical activity. Prev. 1916–2010. J. Urban Health 88, 381–389. https://doi.org/10.1007/s11524-011- Med. (Baltim). 76, 31–36. https://doi.org/10.1016/j.ypmed.2015.04.001. 9585-2. Bancroft, C., Joshi, S., Rundle, A., Hutson, M., Chong, C., Weiss, C.C., Genkinger, J., Glaser, B., 2003. The Grounded Theory Perspective II: Description’s Remodeling of Neckerman, K., Lovasi, G., 2015. Association of proximity and density of parks and Grounded Theory Methodology. Mill Valley, California. objectively measured physical activity in the United States: a systematic review. Soc. Hill, T.D., Ross, C.E., Angel, R.J., 2005. Neighborhood disorder, psychophysiological Sci. Med. 138, 22–30. https://doi.org/10.1016/j.socscimed.2015.05.034. distress, and health. J. Health Soc. Behav. 46, 170–186. https://doi.org/10.1177/ Beenackers, M.A., Kamphuis, C.B.M., Giskes, K., Brug, J., Kunst, A.E., Burdorf, A., van 002214650504600204. Lenthe, F.J., 2012. Socioeconomic inequalities in occupational, leisure-time, and Hinckson, E., Cerin, E., Mavoa, S., Smith, M., Badland, H., Stewart, T., Duncan, S., transport related physical activity among European adults: a systematic review. Int. Schofield, G., 2017. Associations of the perceived and objective neighborhood J. Behav. Nutr. Phys. Activ. 9 https://doi.org/10.1186/1479-5868-9-116. environment with physical activity and sedentary time in New Zealand adolescents. Biering-Sørensen, F., Alexander, M.S., Burns, S., Charlifue, S., Devivo, M., Dietz, V., Int. J. Behav. Nutr. Phys. Activ. 14, 1–15. https://doi.org/10.1186/s12966-017- Krassioukov, A., Marino, R., Noonan, V., Post, M.W.M., Stripling, T., Vogel, L., 0597-5. Wing, P., 2011. Recommendations for translation and reliability testing of Hoffimann, E., Barros, H., Ribeiro, A.I., 2017. Socioeconomic inequalities in green space international spinal cord injury data sets. Spinal Cord 49, 357–360. https://doi.org/ quality and Accessibility—evidence from a Southern European city. Int. J. Environ. 10.1038/sc.2010.153. Res. Publ. Health 14. https://doi.org/10.3390/ijerph14080916. Bilal, U., Díez, J., Alfayate, S., Gullón, P., del Cura, I., Escobar, F., Sandín, M., Franco, M., Jennings, V., Bamkole, O., 2019. The relationship between social cohesion and urban 2016. Population cardiovascular health and urban environments: the Heart Healthy green space: an avenue for health promotion. Int. J. Environ. Res. Publ. Health 16. Hoods exploratory study in Madrid, Spain. BMC Med. Res. Methodol. 16, 104. https://doi.org/10.3390/ijerph16030452. https://doi.org/10.1186/s12874-016-0213-4. Jones, A., Hillsdon, M., Coombes, E., 2009. Greenspace access, use, and physical activity: Brownson, R.C., Hoehner, C.M., Day, K., Forsyth, A., Sallis, J.F., 2009. Measuring the understanding the effects of area deprivation. Prev. Med. (Baltim). 49, 500–505. built environment for physical activity. Am. J. Prev. Med. 36, S99–S123. https://doi. https://doi.org/10.1016/j.ypmed.2009.10.012. org/10.1016/j.amepre.2009.01.005 e12. Kabisch, N., Korn, H., Stadler, J., 2017. Nature-Based Solutions to Climate Change Bryman, A., 2006. Integrating quantitative and qualitative research: how is it done? Adaptation in Urban Areas. Springer International Publishing, Cham. Qual. Res. 6, 97–113. https://doi.org/10.1177/1468794106058877. Klompmaker, J.O., Janssen, N.A.H., Bloemsma, L.D., Gehring, U., Wijga, A.H., van den Cavaleiro Rufo, J., Ribeiro, A.I., Paciência, I., Delgado, L., Moreira, A., 2020. The Brink, C., Lebret, E., Brunekreef, B., Hoek, G., 2019. Residential surrounding green, influence of species richness in primary school surroundings on children lung air pollution, traffic noise and self-perceived general health. Environ. Res. 179, function and allergic disease development. Pediatr. Allergy Immunol. 31, 358–363. 108751 https://doi.org/10.1016/j.envres.2019.108751. https://doi.org/10.1111/pai.13213. 9 M. Fontán-Vela et al. Health and Place 71 (2021) 102660 Knapp, M., Gustat, J., Darensbourg, R., Myers, L., Johnson, C., 2019. The relationships Quillian, L., Pager, D., 2001. Black neighbors, higher crime? The role of racial between park quality, park usage, and levels of physical activity in low-income, stereotypes in evaluations of neighborhood crime. Am. J. Sociol. 107, 717–767. African American neighborhoods. Int. J. Environ. Res. Publ. Health 16. https://doi. https://doi.org/10.1086/338938. org/10.3390/ijerph16010085. Rigolon, A., 2016. A complex landscape of inequity in access to urban parks: a literature Knobel, P., Kondo, M., Maneja, R., Zhao, Y., Dadvand, P., Schinasi, L.H., 2021. review. Landsc. Urban Plann. 153, 160–169. https://doi.org/10.1016/j. Associations of objective and perceived greenness measures with cardiovascular risk landurbplan.2016.05.017. factors in Philadelphia, PA: a spatial analysis. Environ. Res. 197, 110990 https://doi. Rivera Navarro, J., Franco Tejero, M., Conde Espejo, P., Sandín Vázquez, M., Gutiérrez org/10.1016/j.envres.2021.110990. Sastre, M., Cebrecos, A., Sainz Muñoz, A., Gittelsohn, J., 2019. Understanding Urban Kothencz, G., Blaschke, T., 2017. Urban parks: visitors’ perceptions versus spatial Health Inequalities: Methods and Design of the Heart Health Hoods Qualitative indicators. Land Use Pol. 64, 233–244. https://doi.org/10.1016/j. Project. Gac. Sanit. 33, 517–522. https://doi.org/10.1016/j.gaceta.2018.07.010. landusepol.2017.02.012. Rose, G., 1985. Sick Individuals and Sick Populations, vol. 14, pp. 32–38. Langemeyer, J., Connolly, J.J.T., 2020. Weaving notions of justice into urban ecosystem Sampson, R.J., Laub, J.H., 2003. Life-course desisters? Trajectories of crime among services research and practice. Environ. Sci. Pol. 109, 1–14. https://doi.org/ delinquent boys followed to age 70. Criminology 41, 555–592. https://doi.org/ 10.1016/j.envsci.2020.03.021. 10.1111/j.1745-9125.2003.tb00997.x. Lapham, S.C., Cohen, D.A., Han, B., Williamson, S., Evenson, K.R., McKenzie, T.L., Santos, M.P.M., Rech, C.R., Alberico, C.O., Fermino, R.C., Rios, A.P., David, J., Reis, R.S., Hillier, A., Ward, P., 2016. How important is perception of safety to park use? A four- Sarmiento, O.L., McKenzie, T.L., Mota, J., 2016. Utility and reliability of an app for city survey. Urban Stud. 53, 2624–2636. https://doi.org/10.1177/ the system for observing play and recreation in Communities (iSOPARC®). Meas. 0042098015592822. Phys. Educ. Exerc. Sci. 20, 93–98. https://doi.org/10.1080/ Leal, J., Sorando, D., 2015. Economic crisis, social change and segregation processes in 1091367X.2015.1120733. Madrid. In: Tammaru, T., van Ham, M., Marcinczak, S., Musterd, S. (Eds.), Socio- Schüle, S.A., Hilz, L.K., Dreger, S., Bolte, G., 2019. Social inequalities in environmental Economic Segregation in European Capital Cities. Taylor & Francis Group, London, resources of green and blue spaces: a review of evidence in the WHO European p. 414. https://doi.org/10.4324/9781315758879. region. Int. J. Environ. Res. Publ. Health 16. https://doi.org/10.3390/ Leslie, E., Cerin, E., Kremer, P., 2010. Perceived neighborhood environment and park use ijerph16071216. as mediators of the effect of area socio-economic status on walking behaviors. Seiluri, T., Lahti, J., Rahkonen, O., Lahelma, E., Lallukka, T., 2011. Changes in J. Phys. Activ. Health 7, 802–810. https://doi.org/10.1123/jpah.7.6.802. occupational class differences in leisure-time physical activity: a follow-up study. Int. Ma, L., Dill, J., Mohr, C., 2014. The objective versus the perceived environment: what J. Behav. Nutr. Phys. Activ. 8, 1–8. https://doi.org/10.1186/1479-5868-8-14. matters for bicycling? Transportation (Amst) 41, 1135–1152. https://doi.org/ Stevenson, M., Thompson, J., de Sá, T.H., Ewing, R., Mohan, D., McClure, R., Roberts, I., 10.1007/s11116-014-9520-y. Tiwari, G., Giles-Corti, B., Sun, X., Wallace, M., Woodcock, J., 2016. Land use, Markevych, I., Schoierer, J., Hartig, T., Chudnovsky, A., Hystad, P., Dzhambov, A.M., de transport, and population health: estimating the health benefits of compact cities. Vries, S., Triguero-Mas, M., Brauer, M., Nieuwenhuijsen, M.J., Lupp, G., Lancet 388, 2925–2935. https://doi.org/10.1016/S0140-6736(16)30067-8. Richardson, E.A., Astell-Burt, T., Dimitrova, D., Feng, X., Sadeh, M., Standl, M., Triguero-Mas, M., Dadvand, P., Cirach, M., Martínez, D., Medina, A., Mompart, A., Heinrich, J., Fuertes, E., 2017. Exploring pathways linking greenspace to health: Basagaña, X., Gražulevičienė, R., Nieuwenhuijsen, M.J., 2015. Natural outdoor Theoretical and methodological guidance. Environ. Res. 158, 301–3017. https://doi. environments and mental and physical health: relationships and mechanisms. org/10.1016/j.envres.2017.06.028. Environ. Int. 77, 35–41. https://doi.org/10.1016/j.envint.2015.01.012. McCormack, G.R., Rock, M., Toohey, A.M., Hignell, D., 2010. Characteristics of urban Uijtdewilligen, L., Waters, C., Aw, S., et al., 2019. The Park prescription study: parks associated with park use and physical activity: a review of qualitative research. Development of a community-based physical activity intervention for a multi-ethnic Health Place 16, 712–726. https://doi.org/10.1016/j.healthplace.2010.03.003. Asian population. PLoS ONE 14 (6), 1–24. https://doi.org/10.1371/journal. Mckenzie, T.L., Cohen, D.A., 2006. SOPARC (System for Observing Play and Recreation pone.0218247. in Communities) Description and Procedures Manual SOPARC (System for Observing Van Welsem, J., Wittebrood, K., De Graaf, N.D., 2006. Socioeconomic dynamics of Play and Recreation in Communities) PURPOSE. neighborhoods and the risk of crime victimization: a multilevel study of improving, Nieuwenhuijsen, M.J., Khreis, H., Triguero-Mas, M., Gascon, M., Dadvand, P., 2017. Fifty declining, and stable areas in The Netherlands. Soc. Probl. 53, 226–247. https://doi. shades of green. Epidemiology 28, 63–71. https://doi.org/10.1097/ org/10.1525/sp.2006.53.2.226. EDE.0000000000000549. Vaughan, C.A., Colabianchi, N., Hunter, G.P., Beckman, R., Dubowitz, T., 2018. Park use Orstad, S.L., McDonough, M.H., Stapleton, S., Altincekic, C., Troped, P.J., 2017. in low-income urban neighborhoods: who uses the parks and why? J. Urban Health A systematic review of agreement between perceived and objective neighborhood 95, 222–231. https://doi.org/10.1007/s11524-017-0221-7. environment measures and associations with physical activity outcomes. Environ. Warburton, D.E.R., Bredin, S.S.D., 2017. Health benefits of physical activity: a systematic Behav. 49, 904–932. https://doi.org/10.1177/0013916516670982. review of current systematic reviews. Curr. Opin. Cardiol. 32, 541–556. https://doi. Oscilowicz, E., Honey-Rosés, J., Anguelovski, I., Triguero-Mas, M., Cole, H., 2020. Young org/10.1097/HCO.0000000000000437. families and children in gentrifying neighbourhoods: how gentrification reshapes use Ward, P., McKenzie, T.L., Cohen, D., Evenson, K.R., Golinelli, D., Hillier, A., Lapham, S. and perception of green play spaces. Local Environ. 25, 765–786. https://doi.org/ C., Williamson, S., 2014. Physical activity surveillance in parks using direct 10.1080/13549839.2020.1835849. observation. Prev. Chronic Dis. 11, 1–9. https://doi.org/10.5888/pcd10.130147. O’Donoghue, G., Kennedy, A., Puggina, A., Aleksovska, K., Buck, C., Burns, C., Wen, M., Zhang, X., Harris, C.D., Holt, J.B., Croft, J.B., 2013. Spatial disparities in the Cardon, G., Carlin, A., Ciarapica, D., Colotto, M., Condello, G., Coppinger, T., distribution of parks and green spaces in the USA. Ann. Behav. Med. 45, 18–27. Cortis, C., D’Haese, S., De Craemer, M., Blasio, A. Di, Hansen, S., Iacoviello, L., https://doi.org/10.1007/s12160-012-9426-x. Issartel, J., Izzicupo, P., Jaeschke, L., Kanning, M., Ling, F., Luzak, A., Napolitano, G., WHO Regional Office for Europe, 2016. Urban Green Spaces and Health. Copenhagen. Nazare, J.A., Perchoux, C., Pesce, C., Pischon, T., Polito, A., Sannella, A., Schulz, H., Wilbur, J., Chandler, P., Dancy, B., Choi, J., Plonczynski, D., 2002. Environmental, Simon, C., Sohun, R., Steinbrecher, A., Schlicht, W., MacDonncha, C., Capranica, L., policy, and cultural factors related to physical activity in urban, african American Boccia, S., 2018. Socio-economic determinants of physical activity across the life women. Women Health 36, 17–28. https://doi.org/10.1300/J013v36n02_02. course: a “DEterminants of DIet and Physical ACtivity” (DEDIPAC) umbrella Willey, J.Z., Paik, M.C., Sacco, R., Elkind, M.S.V., Boden-Albala, B., 2010. Social literature review. PloS One 13, 1–24. https://doi.org/10.1371/journal. determinants of physical inactivity in the NOrthern Manhattan Study (NOMAS). pone.0190737. J. Community Health 35, 602–608. https://doi.org/10.1007/s10900-010-9249-2. Papas, M.A., Alberg, A.J., Ewing, R., Helzlsouer, K.J., Gary, T.L., Klassen, A.C., 2007. The Wilson, D.K., Kirtland, K.A., Ainsworth, B.E., Addy, C.L., 2004. Socioeconomic status and built environment and obesity. Epidemiol. Rev. 29, 129–143. https://doi.org/ perceptions of access and safety for physical activity. Ann. Behav. Med. 28, 20–28. 10.1093/epirev/mxm009. https://doi.org/10.1207/s15324796abm2801_4. Pearce, J., Cherrie, M., Shortt, N., et al., 2018. Life course of place: A longitudinal study Young, D., Saksvig, B.I., Wu, T.T., Zook, K., Li, X., Champaloux, S., Grieser, M., Lee, S., of mental health and place. Transactions of the Institute of British Geographers 43 Treuth, M.S., 2014. Multilevel correlates of physical activity for early, mid, and late (4), 555–572. https://doi.org/10.1111/tran.12246. adolescent girls. J. Phys. Activ. Health 11, 950–960. https://doi.org/10.1123/ Prüss-Ustün, A., Wolf, J., Corvalán, C., Bos, R., Neira, M., 2016. Preventing disease jpah.2012-0192. through healthy environments: a global assessment of the environmental burden of disease. Toxicol. Lett. https://doi.org/10.1016/j.toxlet.2016.07.028. 10