What Is the Relationship between the Neighbourhood Built Environment and Time Spent in Outdoor Play? A Systematic Review

Outdoor play has been associated with children’s and adolescents’ healthy development and physical activity. Attributes of the neighbourhood built environment can influence play behaviours. This systematic review examined the relationship between attributes of the neighbourhood built environment and the time children and adolescents (0–18 years) spend in self-directed outdoor play. We identified and evaluated 18 relevant papers using the Mixed Methods Appraisal Tool and developed a narrative synthesis of study results. We found moderate evidence that lower traffic volumes (ages 6–11), yard access (ages 3–10), and increased neighbourhood greenness (ages 2–15) were positively associated with time spent in outdoor play, as well as limited evidence that specific traffic-calming street features such as fewer intersections, low traffic speeds, neighbourhood disorder, and low residential density were positively associated with time spent in outdoor play. To our knowledge, this is the first systematic review on this topic. The limited number of “good quality” studies identified highlights the need for additional research on the topic.


Introduction
Play is a central activity in children's lives around the world [1]. Adults reflecting on favourite play memories often recall outdoor play, particularly in natural settings, remembering opportunities for freedom, fun, creativity, and skill-and confidence-building [2][3][4]. Recent research has identified a myriad benefits for children's development, health and well-being. One of the most robust findings links time spent in outdoor play to physical activity, indicating that children playing outside are more physically active and less sedentary than when indoors [5,6]. A study of 10-13-year-old children and adolescents found that outdoor play contributed 36 min/day to physical activity, versus 40 min/day for organized sports, 17 min/day for active travel and 26 min/day for curriculum based physical activity [7]. Outdoor play has further been associated with motor, visual, and cognitive development, socio-emotional learning and mental health [8][9][10][11]. The benefits of outdoor play have also been recognized by pediatricians and public health professionals, who have stressed the importance of daily opportunities for outdoor play [12,13].
In this review, the independent variable is the "neighbourhood built environment attribute", defined as a characteristic of the physical outdoor space near the residence of child or adolescent study participants, including yards, streets and public open space. We included front and back yards in the neighbourhood play environment because of their role as an intermediary between the street, the alley, and the home. Due to our interest in understanding the influence of neighbourhood planning and design, we focused on environmental characteristics, such as yard size and intersection density, as opposed to equipment such as slides and swings. We included pedestrian infrastructure such as benches, water fountains, and garbage cans because of their impact on street width. We included school grounds if they were examined outside of school hours, independently of programs such as after-school care. Because parent and child perceptions of the environment are important predictors of play-related behaviours such as child independent mobility and physical activity, we included both subjectively-perceived and objectively-measured attributes of the built environment.
Play is defined as "freely chosen, personally directed, intrinsically motivated behaviour that actively engages the child" [40]. We only included studies that examined "outdoor" play because of the numerous health and developmental benefits associated with outdoor play [16], as well as to investigate the relationship between play and the built environment. To ensure the inclusion of papers studying outdoor play behaviours in adolescents, we included the terms "hanging out", "unstructured time" and "leisure" in our searches, the most frequent terms encountered in our survey of the literature. We decided that use of the term "play" was a sufficient descriptor of play behaviours in participants, even if the term is rarely defined in publications. Moreover, we did not include independent mobility as a type of play. Though studies have demonstrated a positive link between independent mobility and neighbourhood outdoor play [41,42], independent mobility behaviours can also include travel behaviours unrelated to play. Finally, we decided that the primary outcome of selected studies should be "time" spent in outdoor play. This criterion allowed us to identify environmental features with a measurable impact on play behaviour.
In summary, we included studies in the review if (a) they included children and adolescents aged 0-18, if (b) they reported a subjective or objective outdoor neighbourhood built environment characteristic as an independent variable, and if (c) this variable was linked to time spent in self-directed outdoor play.

Study Exclusion Criteria
Articles were excluded if they studied the link between outdoor play and non-neighbourhood environments, such as indoor environments. Outcome behaviours were not considered outdoor play if the activities described were explicitly directed by adults, including organized sports, time spent in school, childcare, or an after-school program. Studies which examined the link between the neighbourhood built environment and outdoor play without including a measure of time or duration were also excluded.

Search Strategy
The neighbourhood play search strategy is described in Appendix A. Because of the interdisciplinary scope of the topic, we searched six electronic databases during the review process: the Avery Index, MEDLINE (Ovid), PsycInfo (EBSCO), SPORTDiscus (EBSCO), ERIC (EBSCO), CINAHL (EBSCO). Additionally, we manually searched the Journal of Children, Youth and Environments because it is a journal that has traditionally published research relevant to this topic. All databases were searched on 28 January 2019. Finally, a manual search of the reference list of identified studies was conducted to search for potential additional studies, as suggested by the Cochrane Handbook for Systematic Reviews of Interventions [43]. All studies identified in the manual search were found to be duplicates of studies identified by the database search.
Search terms included Boolean combinations of the following words: play, leisure, hanging out, unstructured time, built environment, physical environment, neighbourhood, street, yard, child, youth, adolescent, teen, time and duration (Appendix A). Two independent reviewers examined the titles and abstracts of all articles identified by the search strategy and read all selected full text articles. Any discrepancies were resolved by discussion and consensus, including a third reviewer when necessary. Consensus was reached on all decisions of study eligibility.

Data Extraction and Quality Assessment
Data extraction was completed by one researcher (AL) and checked by another (JV). The strength of association between dependent and independent variables was extracted for each selected study, and methodological quality of studies was assessed by one researcher (JV). Another researcher (MB) reviewed all extracted effect magnitudes and checked a subset of the methodological quality assessment. When data appeared to be incomplete or incorrect, corresponding authors were contacted by email for additional information. Data extraction identified all study methods and results related to the inclusion criteria, and transcribed these in Tables 1-4, using each study's exact wording. We selected the Mixed Methods Appraisal Tool (MMAT) to evaluate the methodological quality of included studies because it allowed for a reliable and efficient analysis of quantitative-descriptive research. The tool evaluated the following criteria: participant recruitment (if the study reports how representative the sample is of the population), outcome measurements (whether outcome measures have been standardized or validated), participant controls (if demographic characteristics are reported and if the most important factors are controlled for in analyses), and response rates (>80% complete data and >60% response rate). For each criterion, the study received one star if it was reported by the study and met (yes), and zero stars if it was not reported by the study (can't tell) or was reported, but not met (no). Studies were graded according to the number of criteria met (* = 1 criterion met, with a maximum of **** = 4 criteria met). Details on MMAT methodology are described elsewhere [44].    Parent questionnaire (identical at child age 5 and 7 years): Play defined as: total duration of unstructured outside play in an average week, without organized sports, school physical education, and active transport. First, parents asked to report number of days their child played outside in an average week for the last four week. Second, parent asked to report the average duration of outside play. Parent questionnaire: "report how often their child played in the yard at home, their own street/court/footpath, and the park/playground outside school hours on weekdays and on weekend days during a typical week." "Responses to weekday items were marked on a five-point scale ranging from never/rarely to five days per week; and for the weekend items, on a six-point scale ranging from never/rarely to every Saturday and Sunday." ("Instrument to assess children's outdoor play in various locations" [ Notes: a A 'neighbourhood size' measure was used by several studies to define an area for analysis which reflected participants' neighbourhood environments. b The term 'urban' was used by several studies to describe participants' neighbourhood type. Other studies compared a variety of neighbourhood types. When neighbourhood type was unspecified, this is marked by (-).     Grigsby-Toussaint [47] Increase in neighbourhood greenness (NDVI) (objective: index) Total outdoor PA in minutes per day (sum of child's total "quiet" and "active" OP time) (subjective: parent report) a

Analysis
Meta-analysis was planned for sufficiently homogeneous data with respect to statistical and methodological characteristics. Otherwise, narrative syntheses of research outcomes were conducted to highlight patterns in study methodologies and results. We chose to summarize study results as per Tompa, Trevithick and McLeod's [58] best evidence synthesis guidelines, as per Appendix B. This method takes into account the number, the quality and the consistency of studies to evaluate the strength of the evidence on a topic [58]. According to the methodology, all medium and high quality studies are included in the narrative analysis, in order to highlight the most reliable results. Low quality studies were not included in the analysis. The results of the synthesis provide a general sense of which neighbourhood features are the most influential: often studies had different results for different age groups and genders. We used MMAT assessments for the quality component of the synthesis (one and two MMAT stars = "low quality study", 3 stars = "medium quality study", 4 stars = "high quality study").

Results
From an initial result of 2876 papers, we identified 18 articles that fit our criteria, from 18 different studies, published between 2004 and 2016 ( Figure 1).

Analysis
Meta-analysis was planned for sufficiently homogeneous data with respect to statistical and methodological characteristics. Otherwise, narrative syntheses of research outcomes were conducted to highlight patterns in study methodologies and results. We chose to summarize study results as per Tompa, Trevithick and McLeod's [58] best evidence synthesis guidelines, as per Appendix B. This method takes into account the number, the quality and the consistency of studies to evaluate the strength of the evidence on a topic [58]. According to the methodology, all medium and high quality studies are included in the narrative analysis, in order to highlight the most reliable results. Low quality studies were not included in the analysis. The results of the synthesis provide a general sense of which neighbourhood features are the most influential: often studies had different results for different age groups and genders. We used MMAT assessments for the quality component of the synthesis (one and two MMAT stars = "low quality study", 3 stars = "medium quality study", 4 stars = "high quality study").

Results
From an initial result of 2876 papers, we identified 18 articles that fit our criteria, from 18 different studies, published between 2004 and 2016 ( Figure 1).

Study Characteristics
A variety of instruments were used to measure time spent in outdoor play (see Table 2

Study Results
The selected articles identified several associations between the neighbourhood built environment and time spent in outdoor play, with significant results summarized in Table 3. Heterogeneity in measurement of time spent in outdoor play did not allow for meta-analysis. The results of the best evidence synthesis are summarized in Table 4. In order to synthesize findings, we grouped similar study results into eight themes and 17 attributes (subthemes) of the built environment. The themes and subthemes were developed inductively through consideration of the characteristics of the described features and labeled accordingly. For example, in subtheme "traffic volume", we included Bringolf-Isler's measure of parents' perception of a "problem to play outdoors because of traffic" [36], Page's measure of parents' perception of "traffic safety"(a compound measure including the variable "heavy traffic") [41], Aarts' audits of the "presence of home zones" [45], and Lee's measure of street segments with "low traffic volumes" [51]. This organization of results required certain theoretical assumptions, such as "home zones" being areas where traffic volumes would be low. Authors were contacted when terms were unclear. In Table 4, studies with an MMAT methodological rating below ***("medium quality") are in grey font. Their results are not included in the best evidence synthesis [see Appendix B for summary of guidelines].

Public Open Space Characteristics (9 Studies)
This theme summarizes the impact of open space that is distinct from the home (as opposed to the yard). The review found moderate evidence that public open spaces had no effect on play: five medium/high-quality studies found no relationship between public open space attributes and time spent in outdoor play. However, other studies did find associations: in one medium-quality study, a higher proportion of parks, woods, and agriculture per 2.5 hectares around the home predicted more time spent in outdoor play for children aged 6 to 10 [36]. Inversely, one high-quality study associated a higher number of formal outdoor play spaces (play grounds, school yards, paved play grounds, and half pipe or skating track) with less time spent in play in the 7-12 age range [45]. The authors suggest that this result may be linked to the "street play" culture of the Netherlands, as well as to the quality of surrounding play spaces: play spaces perceived as unsafe may impede outdoor play.

Street Characteristics (8 Studies)
This theme summarizes attributes related to street proportions and street infrastructure. The review found limited evidence that street characteristics have an influence on outdoor play: traffic calming features and pedestrian amenities appear to have a positive impact, while walkability and intersections seem to have a negative impact. Sidewalks, traffic lights, speed bumps, home zones (woonerven in the Netherlands: street configurations which often have speed limits of 15km/hr), roundabouts and "safe places to cross" were associated with more time spent in outdoor play in two high-and medium-quality studies [41,45]. Aarts et al. [45] found that pedestrian crossings had mixed results, while street lighting and safety islands were associated with less outdoor play. The presence of parallel parking spaces and parking lots had a positive association with outdoor play for older boys in Aarts et al. [45], who suggest that parallel parking spots can provide buffers between the street and the sidewalk/yard play area, while Lee et al. [51] found a negative association for both genders 6-11 for parking spaces, as part of a compound measure of "path obstructions". In Lee et al. [51], pedestrian amenities such as benches and water fountains were associated with more outdoor play. Greater intersection density was associated with less time spent in outdoor play for children in one medium-quality study [51], and another high-quality study identified that the "presence of intersections" had the same association [45]. In one high-quality study, living in a "walkable neighbourhood" was linked to more outdoor play in a park, but to less outdoor play in a cul-de-sac or driveway [50]. Lee et al. [51] also found that walkability, defined as a combination of land use, street connectivity and residential density, was associated with less outdoor play.

Traffic Characteristics (7 Studies)
The review found moderate evidence that low traffic volumes have a positive impact on outdoor play. Three high-and medium-quality studies reported that low traffic volumes were associated with more time spent in outdoor play [41,45,51], and one study found that traffic was a barrier to outdoor play [36]. Other subgroups were not affected: Bringolf-Isler et al. [36] found that traffic volume had no effect on the outdoor play of 13-14 year olds, Page et al. [41] had the same result for boys 10-11, as did Aarts et al. [45] for girls 4-6 and 7-12 year olds of both genders. Aarts et al. [45] found that home zones were linked to increased outdoor play, while the presence of 30 km/hr zones were associated with less outdoor play, and found no effect on girls aged 4-12.

Housing Characteristics (7 Studies)
This theme summarizes the effects of living in a specific type of housing, living while surrounded by a certain type of housing, or living in a certain population density, a characteristic closely related to building type. The review found limited evidence linking housing and outdoor play: both higher density and detached homes were associated with less play, while duplexes had the opposite effect. Kimbro et al. [37] found that living in public housing was linked to more outdoor play in five year olds. Aarts et al. [35] found that girls 4-6 living in a detached residence spent less time in outdoor play, while boys 4-6 living in a rental property or a duplex spent more time in outdoor play. Kimbro et al. [37] and Aarts et al. [35] (girls 4-6, boys [10][11][12] found that living in an apartment was linked to less outdoor play. Greater neighbourhood population density was also linked to less time spent in outdoor play in two medium-quality studies for 6-11 year olds [36,51], though Lee et al. [51] examined residential density within a measure of neighbourhood walkability scores, wherein greater density is linked to greater walkability. Bringolf-Isler et al. [36] found that neither population or housing density had an effect on the outdoor play of 13-14 year olds.

Yard Characteristics (6 Studies)
This theme includes all private outdoor space surrounding the home. The review found moderate evidence that yard access was positively associated with more outdoor play. One high-quality and two medium-quality studies reported that access to a yard was associated with more time spent in outdoor play [52], or that the absence of yards was linked to less time spent in outdoor play [35,36]. Aarts et al. [35] reported that the absence of a yard predicted more outdoor play in girls 4 to 6, while two studies found no relationship between yard access and outdoor play [35,36], Bringolf-Isler et al. [36] finding no effect on 13-14 year olds. One medium quality study found that yard size had no effect [49].

Neighbourhood Greenness (4 Studies)
This theme includes measures that refer to general neighbourhood "greenness", as opposed to a specific place such as a park. Studies identified greenness through satellite imagery [47], features such as street trees, flower beds and recreational areas [48] or governmental designation as a neighbourhood type with extensive greenery [35]. The review found moderate evidence that neighbourhood greenness was a predictor of more time spent in outdoor play, as identified in three high-and medium-quality studies [35,48,53]. One longitudinal study found that a perceived increase of "nature" in the neighbourhood was linked to less time spent in walking (for leisure) for boys 12-15 [48].

Physical Disorder (4 Studies)
This theme includes measures of house maintenance, litter, graffiti, vandalism and dog feces. The review found mixed evidence of its effect on outdoor play. One medium-quality study linked greater physical disorder around the home with more outdoor play in five year-olds [37], while one high-quality study found that better house maintenance was associated with boys 10-12 spending less time in outdoor play [45]. Both studies suggest that increased physical disorder in a neighbourhood can support play. Two medium-quality studies found no association between disorder and time spent in outdoor play [54].
Other built environment attributes significantly associated with outdoor play in the 18 selected studies include the presence of water, access to a diversity of travel routes and living in a city centre [35]. Because these results were not mentioned in more than one selected study, they are not included here (see Table 3).

Summary of Results
The 18 studies that met our inclusion criteria examined a broad range of neighbourhood built environment features. In most studies, play behaviours were subjectively reported by parents, thus results depended on their individual interpretation of the term "play" (see Table 2). We did not identify any publications with objective measures of neighbourhood play. According to the MMAT methodological appraisal, six studies met the criteria for low methodological quality, eight studies met medium quality criteria and four studies met high quality criteria (see Table 3).
Several studies identified gender-specific results: concerns about traffic safety seem more likely to constrain girls' play, and pedestrian infrastructure was more often linked to increased play in girls. Inversely, certain traffic-related features, such as parking, were linked to more time spent in outdoor play for boys, aged 7 and up. This may be related to boys' use of paved surfaces for team games or skateboarding. In this review, adolescents' outdoor play was less likely to be associated with built environment features. Intersections, traffic volume and building/population density had no effect on participants aged 12 and up. We also identified regional trends: neighbourhood greenness, public open space, and housing density were more likely to be studied in Europe, whereas walkability was studied exclusively in North America.

Play and Urban Design
At first glance, the features of a "playable" neighbourhood identified in this review describe a rural or suburban neighbourhood, with private yards, extensive greenery and limited traffic. Suburban typologies, developed in the post-WWII era, emphasize the "nuclear family" and private spaces for children's play [74]. Recently, extensive research on low-density neighbourhoods has shown their negative impacts on adult health by reducing active travel behaviours [75] and requiring a high consumption of fossil fuels [76], highlighting that though suburban characteristics may support child and adolescent outdoor play, they are also associated with disadvantages for other age groups.
Most current trends in urban design (such as "smart growth" and new urbanism) promote dense typologies which emphasize active travel and public transit [76]. However, this review highlights evidence that density and certain accompanying features can have a negative impact on children's outdoor play. One approach to this dilemma may be to identify denser neighbourhood typologies which include many, if not all the playable features synthesized in this review. For example, compact urban neighbourhoods with traffic calming features, such as cul-de-sacs that remain permeable to cyclists and pedestrians, 15 km/hr speed limits, and features, such as benches and trees, that encourage chance encounters and conversation. However, more research on small-scale street features is required to determine their impact. For example, garbage cans could be perceived as either a pedestrian amenity or a path obstruction, as in Lee et al. [51]. Ready access to parks may be less important to outdoor play: this review identified that proximate green spaces such as yards and general neighbourhood greenness have a greater impact on outdoor play than public open space. This review found that apartments and detached homes were negatively associated, while duplexes were positively associated with time spent in outdoor play: providing shared outdoor space, or other "doorstep" play spaces may be one strategy for capitalizing on the positive effects and mitigating the negative effects of housing density on children's outdoor play.

Outdoor Play and Physical Activity
Several existing systematic reviews examine the relationship between the built environment and children's physical activity. A 2006 review found that children's physical activity was positively related to the proximity of recreational facilities and schools, as well as to the presence of sidewalks, controlled intersections, accessible destinations, and public transportation [77]. Physical activity was negatively associated with a high number of roads, high traffic density, high traffic speeds, area deprivation, and perceptions of local crime. The present review found similar results, though two studies in this review found that physical disorder had a positive association with play, especially for boys [37,45]. A 2015 review [39] found surprising results concerning objectively-measured built environment attributes and children and youth's physical activity. Elements that attempt to promote play (play facilities, playgrounds, parks, beaches, sports venues, recreational facilities, gyms) were linked to less physical activity in young girls, and pedestrian infrastructure (sidewalks, walking tracks, path lighting, traffic lights, high connectivity streets, local destinations) was linked to less physical activity in children of both genders. Inversely, the combined effect of play infrastructure and walking infrastructure was positively correlated to adolescent girls' and boys' physical activity. We found similar results. Five high-and medium-quality studies in the present review examined links between access to public open space and time spent in outdoor play, with mixed results: four studies found no significant association between the two, and one study found that access to a formal outdoor play facility was negatively associated with outdoor play. We also found that pedestrian infrastructure was positively linked to girls' outdoor play (ages 4-12).

Subjective and Objective Results
Four subcategories in the best evidence synthesis show trends related to their method of assessment of the built environment measurement. Four of the five medium/high quality studies that found no link between public open space and outdoor play measured access to public open space subjectively, i.e., by asking parents or children if parks are located near their home. This suggests that parents' and children's perceptions of public open space access do not influence outdoor play. More studies that objectively measure access to public open space would be an important counterpart to these findings. All three medium/high quality studies that found a link between yard access and outdoor play also used subjective methods. Different housing typologies likely have different yard types (e.g., townhouse courtyards), and parent and child perception of these yards could be important for outdoor play. Inversely, all medium/high quality studies measuring the impact of intersections and neighbourhood greenness used objective measures. This suggests that increasing neighbourhood greenness could lead directly to increased outdoor play, which may be useful for municipal planning strategies.

Limitations
This study is limited by its focus on the built environment: 11 studies out of 18 examined both social and environmental characteristics and found that social factors had a greater or equal effect on outdoor play [35][36][37]41,45,47,50,53,56]. These results emphasize the importance of employing an ecological framework that examines social environments, individual and family characteristics, built environments, and their interdependence when examining the modifiers of play [78]. For example, the results associated with housing type in this review may be associated with household socioeconomic status, which may independently influence outdoor play [79]. Furthermore, built environment interventions may have different effects in different social environments [14]. Increasing accessible green space may not be effective in a neighbourhood where residents do not feel safe allowing children to play outside. Likewise, the interdependence of environmental characteristics must also be considered. For example, increasing greenery without reducing traffic volumes in a neighbourhood may have little effect on outdoor play. Second, a large diversity of methods and measures may have contributed to our mixed results: each study measured neighbourhood environments differently, and many examined a variety of outdoor play behaviours. It is also likely that different countries and communities identify built environment features in disparate ways: a Dutch woonerf, a traffic-calmed residential street, is both similar to and radically different from a Canadian cul-de-sac. The third limitation of this study lies in its analysis of compound built environment measures which combine many variables, such as "attractiveness", "walkability", "poor quality of action-space", "traffic safety", "accessibility" and "pedestrian amenities". Because these compound measures did not report the impact of individual variables, it can be misleading to analyze their association with outdoor play. For example, in Page et al.'s [41] measure of traffic safety, it is possible that only "heavy traffic" is significant, and that "perception of safe places to cross, roads, pollution" have no impact on outdoor play. The fourth limitation of this study is the low number of adolescent participants included. Despite our inclusion of adolescent-specific play terms, we found only four articles examining the built environment and adolescent play behaviours. Fifth, this review is limited by the small number of high quality studies identified. No strong evidence was identified in the evidence synthesis. Sixth, the unstructured outdoor play of adolescents (aged 13-18 years) may differ conceptually from younger children (0-12 years). Only one study that included adolescents divided the results in applicable age groups [48], thus we were unable to meaningfully examine potential differences. Finally, we acknowledge that most of the studies in this review were conducted in high income countries, and as such the results may not reflect realities in other economic or sociocultural contexts.

Conclusions
The findings of this systematic review suggest that the environments where children and adolescents live have some associations with the time they spend in self-directed outdoor play. In this review, we found moderate evidence that low traffic volumes (ages 6-11), yard access (ages 3-10), and increased neighbourhood greenness (ages 2-15) are associated with the time children spend in outdoor play. Interestingly, we found moderate evidence that public open spaces, such as parks and formal outdoor play facilities, have no association with outdoor play time. Evidence was limited about the impact of other features on outdoor play, but features such as fewer intersections, low residential density, low traffic speeds, living in rental housing, living in public housing and higher physical disorder appeared to support outdoor play. Moreover, limited evidence suggests that street features such as sidewalks, traffic lights, speed bumps, home zones and roundabouts have a significant association with girls' play. To our knowledge, this is the first systematic review to examine the impact of the built environment on children's outdoor play. Through a narrative synthesis, the review identifies common trends in international research: diminished "doorstep" play space, loss of vegetation and increased traffic have important impacts on children's outdoor play in communities around the world. Future systematic reviews should consider the qualitative aspect of the relationship between the environment and children's outdoor play, paying special consideration to the influence of child age and gender, and what characteristics are important to the children themselves. Future research should examine different settings, including rural communities and communities with low socioeconomic status. Possible urban design interventions to improve play opportunities could include having numerous accessible play spaces near the home, increased greenery, and trees on residential streets and traffic calming measures. Moreover, interest in designing healthier communities is growing, with several cities proposing major interventions to improve citizen health [80,81]. Cities appear to have enormous potential for reducing health inequalities when they prioritize the provision of safe places to live, work and play for all citizens [80]. Indeed, marginalized and low-income children often have lesser access to safe play spaces, as well as higher rates of illness and injury [32,82]. The results of this review suggest that providing amenities such as neighbourhood vegetation, numerous proximate play spaces and low-traffic zones are important tools for policy makers and designers to support children's outdoor play, an essential component of child development and health. Search date: 18 September 2016 (su(Play) OR "hanging out" OR "unstructured time") AND ("built environment" OR "physical environment" OR neighborhood OR neighbourhood OR park OR playground OR street OR yard) AND (child* OR youth OR adolescent OR teen) AND (time OR duration) AND PD(2000-2016) AND STYPE("scholarly journals") Results: 67 articles Notes: Search 18 September 2016-18 January 2017, zero results. Search date: 17 February 2018 (su(Play) OR "hanging out" OR "unstructured time" OR "leisure") AND ("built environment" OR "physical environment" OR neighborhood OR neighbourhood OR park OR playground OR street OR yard) AND (child* OR youth OR adolescent OR teen) AND (time OR duration) AND PD(2000-2018) AND STYPE("scholarly journals") Results: 0 additional articles Search date: 28 January 2019 (su(Play) OR "hanging out" OR "unstructured time" OR "leisure") AND ("built environment" OR "physical environment" OR neighborhood OR neighbourhood OR park OR playground OR street OR yard) AND (child* OR youth OR adolescent OR teen) AND (time OR duration) AND STYPE("scholarly journals") Limiter: all articles published up to 28 January 2019.