Examining Sociodemographic Differences in Playground Availability and Quality and Associations with Childhood Obesity

Abstract

Background:

Parks and playgrounds are important environmental components for promoting healthy weight among children. The purposes of this study were to examine disparities in access to playgrounds and playground quality by multiple sociodemographic characteristics and the association of playground access and quality with youth weight status.

Methods:

Objectively measured height and weight were collected for all third to fifth grade youth (n = 13,469) in a southeastern US County. Detailed audits of all park playgrounds (n = 95) were concurrently conducted. Playground quality was measured by the presence of eight playground attributes (e.g., good condition and shade). Spatial (GIS) and statistical (multilevel logistic regression) analyses were used to determine access to playgrounds within a ½ mile network buffer, whether playground access and quality varied by the individual's gender, socioeconomic status (SES), and race/ethnicity, and the association between playground access and quality with youth being a healthy weight vs. overweight or obese.

Results:

Higher SES youth were half as likely to have a playground within ½ mile. No disparities were found for playground quality by gender, SES, or race/ethnicity. In unadjusted models, youth with lower quality playgrounds nearby were more likely to be overweight/obese than children with no playground access. However, there were no significant associations for playground access/quality and weight status after adjusting for sociodemographic variables.

Conclusion:

Future research should continue to monitor disparities in diverse metrics of access to and quality of playgrounds and how these are related to weight status and other outcomes among children of different age, income, and racial/ethnic groups.

Introduction

Childhood obesity rates are disconcertingly high in the United States,¹ and the rapid increase can be partially attributed to an imbalance between energy expenditure and caloric intake among youth.² Promoting physical activity (PA) is an effective solution for preventing obesity among youth; however, national data show that less than half of US children meet the PA recommendation of 60 minutes per day.³

Several studies have reported that levels of PA and obesity among children are significantly related to features of the built environment near the home.^4–6 Specifically, increased access to parks is associated with higher PA participation and healthier weight status among youth.^7–10 For example, Wolch et al. observed youth from age 10 until age 18 and found that those with parkland within a 500 m buffer of their home had body mass index (BMI) measurements that were significantly lower than those without parkland in close proximity.¹¹

Researchers have further noted the importance of specific aspects of park environments, such as the available facilities, quality, and safety.^8,12 One study reported that access to particular park facilities was more important than proximity to a park space when considering the effect on obesity risk.⁷ To that end, playgrounds are an important park feature given their appeal and potential for enabling exercise and active play among children.^7,13 Within parks, children engage in greater levels of PA on playgrounds than in other areas.^13,14 Likewise, one study showed that children living within 1 km of a park playground were five times more likely to be of healthy weight than children without such amenities nearby and that playgrounds were the only park feature significantly associated with youth BMI.⁷ Finally, Quigg et al. found that playgrounds were especially important for obese children, as a higher percentage of obese youth utilized playgrounds for their regular PA than either overweight or normal weight children.¹⁵

Although previous research has shown the importance of playgrounds for decreased obesity levels, fewer studies have focused on the specific features or quality of playgrounds.^14,16–18 For instance, one study found that certain attributes, such as size, equipment, safety, and shading, were associated with increased utilization and PA among children.¹⁹ In addition, the authors reported that renovated playgrounds, compared to unrenovated ones, had greater use by children, as well as more vigorous PA.¹⁶ Nevertheless, few studies have examined the effects of playground attributes and quality as they are related to youth obesity.^7,20 Understanding how playground features and quality are associated with childhood weight status could be useful in designing solutions to combat the growing obesity epidemic.

While past research has examined the relationship between access to playgrounds and increased PA and reduced obesity in youth,^7,8,13 it is also important to note sociodemographic variations in these associations. For example, in the Play Across Boston study, playground safety (as measured by 25 items) was inversely associated with the proportion of black residents, youth in poverty, and residents with no high school degree within the census block group.¹⁷ Likewise, in St. Louis, analyses of 100 publicly accessible playgrounds found that safety and maintenance scores were significantly lower in neighborhoods with a greater percentage of black residents.¹⁸ As well, with respect to income, one study reported that children from low SES areas and those living in neighborhoods with high percentages of minorities were less likely to have access to recreational facilities than other groups.²¹

Similar sociodemographic disparities often exist with respect to the impact of playgrounds on youth activity and obesity. One study showed a significant increase in PA among boys with increased access to playgrounds, while less significant relationships were found for girls.²² However, Besenyi et al. concluded that, although both genders showed increased PA levels with playground access within ½ mile from home, playgrounds had a stronger association with PA among girls.²³ With respect to race/ethnicity, Alexander et al. reported that non-Hispanic Black youth with park access were less likely to be obese than non-Hispanic Black children without park access; however, there was no significance in park access and weight status for non-Hispanic White children.¹⁰ In summary, it is important to examine sociodemographic variations in the availability and influence of parks and playgrounds to better understand differential risk factors for childhood obesity.

Given the aforementioned considerations, the three main objectives of this study were to: (1) examine variations in youth access to playgrounds by several youth sociodemographic variables (gender, SES, race/ethnicity), (2) describe differences in playground quality by youth gender, SES, and race/ethnicity, and (3) explore associations between youth weight status and the accessibility and quality of playgrounds.

Methods

Study Setting

This study occurred in a large county in the southeastern United States, with a 2013 total population of 474,266, of which 77.1% was non-Hispanic White, 18.5% was African American, and 8.5% was Hispanic or Latino.²⁴ In 2013, the median household income of the county was $48,886 and ∼15.0% of residents lived below the federal poverty line.²⁴ Within this county, there were 103 publicly accessible parks, with 72 of these containing at least one playground and a total of 95 playgrounds.

Measures and Data Collection

Youth obesity characteristics

Trained physical education teachers from 51 elementary schools collected and recorded the height and weight for all children in third to fifth grade (n = 14,232) enrolled in public schools in the county school district. These data were collected in 2013, concurrent with when the park environmental audits were conducted (described in the playground access and quality section). Height, weight, date of birth, and date of testing were used to calculate BMI percentiles using standardized protocols for youth from the Centers for Disease Control and Prevention.²⁵ Based on BMI percentiles, children in the sample were categorized as healthy weight (<85th percentile) or overweight/obese (≥85th percentile). Location of residence was also obtained for each child in the sample. Using a geographic information system (GIS; ArcGIS 10.2.2), 94.6% of the youths' home addresses were successfully geocoded at the point address level, which was used as the final sample (n = 13,469). Finally, several sociodemographic variables were obtained for each child from the school district. These included gender, full price vs. free/reduced lunch status, and race/ethnicity, which were categorized as White vs. all other race/ethnicities due to sample size limitations.

Playground access and quality

Parks in the county were enumerated using GIS shapefiles and park lists provided by the main municipalities within the study area. Parks not deemed usable and publicly accessible after an in-person audit were excluded from the shapefiles used for the present analyses. ArcGIS 10.2.2 was used to create a ½ mile network buffer around each child's home and then to determine the number of playgrounds that intersected with these buffers. A ½ mile buffer was selected based on a reasonable distance a child might walk or bike to a park and other recent studies showing an association between park or playground access and youth PA or obesity.^23,26,27

To determine the quality of playgrounds based on their attributes, audits of all eligible parks in the county (n = 103) were conducted in 2013 using the Community Park Audit Tool (CPAT).²⁸ For each playground, the CPAT provided a rating of the presence or absence of eight key attributes that were captured in other park audit tools and deemed important for youths' use of playgrounds: usable (everything necessary for use is present and nothing prevents use), good condition (looks clean and maintained), distinct areas for different age groups, colorful equipment (3+ colors), shade cover for at least some (25%+) of the area, benches in/surrounding the area, fence around area (half or more), and separation from the road.²⁹ Based on the presence of these eight attributes, a quality score was determined for each playground with a scale of 0 to 10. The presence of each playground feature added one point to the quality score, while the primary CPAT facility rating attributes of usable and good condition each contributed two points given their central role in perceived playground quality. Each child was assigned an average playground quality score based on all playgrounds within their buffer. Mean playground quality scores were then dichotomized as either lower (≤8) or higher (>8). This cut point was selected because it divided the park playgrounds in the county into relatively equal groups of high (52.6%) and low (47.4%) quality and also because the quality features included in the audit were deemed fairly standard and the majority should be expected at higher quality playgrounds (i.e., playgrounds missing two or more quality attribute points were considered lower quality than playgrounds with less than two quality points absent).

Analyses

For the first research question, multilevel logistic regression was used to examine whether youth gender (male vs. female), SES (full price vs. free/reduced lunch), and race/ethnicity (White vs. other) were significantly associated with youth having at least one playground (vs. 0 playgrounds) within a ½ mile network buffer. Likewise, for the second research question, multilevel logistic regression was again used to examine differences in youths' mean playground quality (higher vs. lower) according to the same three participant demographic characteristics.

Finally, for our third research question, three categories combining playground access and quality were created: (1) youth without a playground within ½ mile, (2) youth whose mean playground quality was lower (≤8), and (3) youth whose mean playground quality was higher (>8). Multilevel logistic regression was used to examine whether this access/quality variable (using no access to playgrounds as the reference variable) was associated with the youth being overweight/obese rather than a healthy weight. Analyses for all three research questions controlled for child age, gender, race/ethnicity, SES, and total population of the child's block group were conducted using SAS 9.4 and were considered significant at p < 0.05.

Results

Characteristics for youth in the study sample are presented in Table 1. The sample included 6846 males (50.8%) and 6623 females (49.2%) with 62.3% of the overall sample being White and 54.7% paying full price for school lunch. The mean age was 9.7 years [standard deviation (SD) = 0.99] and the mean BMI was in the 64th percentile (SD = 29.78), with 65.4% classified as a healthy weight vs. 34.6% as overweight or obese. Within the full sample, 86.3% of children did not have a playground within a ½ mile network buffer of their home.

Table 1.

Sample Characteristics

Youth characteristics (n = 13,469)	Mean or %
Gender
Male	50.8%
Female	49.2%
Age (years, SD)	9.7 (0.99)
Student lunch status
Full priced	54.7%
Free or reduced price	45.3%
Race/ethnicity
White	62.3%
African American	18.9%
Hispanic	11.5%
Other	7.4%
Weight status
Normal weight (<85th percentile)	65.4%
Overweight or obese (≥85th percentile)	34.6%

Playground characteristics (n = 95)
Playground attributes
Usable	98.9%
Colorful	95.8%
Separated from road	94.7%
Good condition	87.4%
Benches	84.2%
Distinct areas for different age groups	82.1%
Shaded	72.6%
Fence surrounding	32.6%
Mean weighted playground score (SD)	8.35 (1.21)
Playground quality category
Lower quality playground (≤8)	47.4%
Higher quality playground (>8)	52.6%

SD, standard deviation.

Overall, park audits revealed that most of the playground quality attributes were quite common across the 95 playgrounds. As shown in Table 1, a very high percentage of playgrounds were rated as usable (98.9%), colorful (95.8%), and separated from the road (94.7%). The only quality attribute with a relatively low frequency (<70%) was the presence of a fence surrounding at least half of the playground, which was found on only 32.6% of playgrounds. Individual playground total quality scores ranged from 5 to 10, with an average of 8.35 (SD = 1.21). Just less than one half of playgrounds (47.4%) were categorized as lower quality (score of ≤8), while 52.6% were classified as higher quality.

For the first research question, Table 2 shows the percentage of youth in each gender, SES, and race/ethnicity group who had access to a playground within ½ mile. Multilevel logistic regression results revealed that there were no differences in playground access according to gender or race/ethnicity, but that higher SES youth were nearly half as likely to have a playground within ½ mile [odds ratio (OR) = 0.60, confidence interval [95% CI] = 0.50–0.72; Table 2]. For the second research question, which included only those youth with a playground within ½ mile (n = 1844), there were no significant differences in higher vs. lower average playground quality by gender, SES, or race/ethnicity (Table 3).

Table 2.

Differences in Youth Playground Access by Sociodemographic Characteristics

	Playground within ½ mile
Youth characteristics	Yes (n = 1844)	No (n = 11,625)	OR	95% CI
Gender
Female	13.2%	86.8%	1.00
Male	14.1%	85.9%	1.09	0.94–1.26
Student lunch status
Free/reduced price	17.4%	82.6%	1.00
Full price	10.6%	89.4%	0.60^*	0.50–0.72
Race/ethnicity
Other	17.8%	82.2%	1.00
White	11.2%	88.8%	1.02	0.86–1.22

p < 0.05.

CI, confidence interval; OR, odds ratio.

Table 3.

Differences in Youth Average Playground Quality by Sociodemographic Characteristics

	Mean playground quality score
Youth characteristics	Lower (≤8) (n = 506)	Higher (>8) (n = 1338)	OR	95% CI
Gender
Female	51.0%	49.0%	1.00
Male	48.6%	51.4%	0.80	0.52–1.25
Student lunch status
Free/reduced price	54.1%	45.9%	1.00
Full price	43.9%	56.1%	0.96	0.54–1.71
Race/ethnicity
Other	55.2%	44.8%	1.00
White	44.4%	55.6%	1.61	0.86–3.02

Finally, in unadjusted models (including only the playground access/quality variable), compared to those who did not have access to a playground within ½ mile, youth who had lower mean playground quality were 1.2 times more likely to be overweight or obese. However, after adjusting for individual characteristics and total block group population, there were no significant associations between lower or higher mean playground quality (vs. no playground access) and weight status (Table 4).

Table 4.

Association between Playground Access and Quality and Weight Status Among Youth

		Unadjusted		Adjusted^a
Playground access and quality category	N	OR	95% CI	OR	95% CI
No playground	11,625	1.00		1.00
Lower mean playground quality score (≤8)	506	1.25^**	1.06–1.48	1.06	0.90–1.24
Higher mean playground quality score (>8)	1338	0.98	0.82–1.16	0.89	0.76–1.05
Lower mean playground quality score (≤8)	506	1.00		1.00
Higher mean playground quality score (>8)	1338	0.74	0.58	0.82^*	0.66–1.03
No playground	11,625	1.00		1.00
Any playground	1844	1.11	0.98–1.26	0.97	0.86–1.10

Adjusted models control for youth age, gender, race/ethnicity, socioeconomic status, and total population in block group.

p < 0.10.

p < 0.05.

We also conducted two additional analyses that are shown in the latter half of Table 4. First, in comparing the subsample of youth who had higher playground quality (n = 927) vs. those with lower playground quality (n = 917), we found, in adjusted models, that the former group was marginally less likely to be overweight or obese (OR = 0.82, 95% CI = 0.66–1.03, p = 0.08). Second, compared to youth with no playground, those with any playground were not more or less likely to be overweight or obese (OR = 0.97, 95% CI = 0.86–1.10).

Discussion

Parks are key components of the built environment, and previous studies have reported that playgrounds are one of the most important components of parks in relation to youth PA and weight status.^7,13,14 Some research has explored how park availability is moderated by variables such as race/ethnicity and SES at the neighborhood (e.g., census tract) level,^30–32 but fewer studies have considered disparities in access to playgrounds for youth from varying sociodemographic backgrounds.^17,18 Likewise, park researchers have begun to examine how specific facilities and amenities within parks are associated with PA among youth,^12,23,33 but less research has focused on playground access and quality specifically or how these factors are related to youth weight status.⁷ Using a large sample of elementary-aged youth from a southeastern US county, this study explored differences in access to and quality of playgrounds by multiple youth sociodemographic characteristics, as well as how combined metrics related to playground access and quality were associated with youth obesity.

Our first research question analyzed access to playgrounds within ½ mile by gender, SES, and race/ethnicity. Across the full sample, less than one in seven children (13.7%) had at least one playground within this distance. There were no differences in playground access according to gender or race/ethnicity, which is encouraging from an equity and environmental justice perspective. Previous research is mixed on whether playgrounds are more important for boys or girls,^22,23,34 but it is perhaps not surprising that their availability is relatively equal given that the geographical distribution of youth by gender is likely uniform.

Conversely, race/ethnicity and SES form more distinct geographical delineations in many communities, including the study county, and much past research have reported that parks and green space are often unevenly accessible to different population groups.^30–32 However, fewer studies have considered equitable access to playgrounds specifically.^17,18,35 While we did not find any significant differences by race/ethnicity, higher SES youth were less likely to have a playground within ½ mile. Indeed, some studies have reported similar counterintuitive findings related to availability of built environment features based on SES or race/ethnicity such that lower SES and high minority populations have increased availability of neighborhood features (e.g., public parks and sidewalks).^35,36 In the current study county, there exists considerable diversity in urbanicity, with a core city center surrounded by both suburban and rural areas. The main city limits are composed of diverse neighborhoods in terms of SES and race/ethnicity, while also having more park density given the greater population density. In contrast, the suburban areas tend to have more affluence, lower overall population density, and fewer parks. Consequently, this may explain the lesser access to playgrounds for high SES youth and these findings suggest that other similar studies may also wish to account for urbanicity and/or population density in such analyses. Regardless, in the current context, the disproportionately better proximity to playgrounds for lower SES children is encouraging given their free cost to access and use, as well as past research showing greater levels of PA in park playgrounds, which can combat high rates of obesity among socioeconomically disadvantaged youth.^13,37

For our second research question, we observed no differences in playground quality by gender, SES, or race/ethnicity of youth. Overall, playground quality in the study location was quite high, with the average score ranging from 5 to 10 and a mean of 8.35. To our knowledge, few prior studies have created a summative measure of multiple playground attributes to ascertain relationships with PA/obesity or sociodemographic characteristics.^17,18 Many of the playground quality features measured in this study, such as fences, shade, benches, or distinct areas for different age groups, may be more important for younger children (or their accompanying parents),³⁸ but we did not examine disparities in playground quality according to age. In contrast, little research has explored whether specific playground attributes are more valuable for youth from certain gender, cultural, or income backgrounds; to that end, the lack of racial/ethnic or socioeconomic disparities observed in our study is likely appropriate, but more research is warranted in this area to tailor the design of these key facilities to local communities.

The third and final objective of this study examined the combined impact of playground access and quality on youth weight status. Unadjusted models showed that youth with access to lower quality playgrounds on average were significantly more likely to be overweight or obese. However, after controlling for demographic characteristics, there were no significant associations found between playground access/quality and weight status. This differs somewhat from an older national study that reported a negative association between access to recreational facilities and obesity rates among Black youth specifically.¹⁰ In the current study, the decent quality of most playgrounds and the limited number of youth in both the lower and higher playground quality groups (about 7% of the total sample in each category) may have limited power to detect differences in youth weight status across playground access and quality classifications.

Limitations

This study had several other limitations that warrant further discussion. First, while the study included a large sample of youth for the study population and location, it was limited to children enrolled in public schools and the narrow geographic area potentially limits the generalizability of the findings to youth in similar cities/counties. As well, this is one of the first studies to assess detailed playground quality attributes, but other elements (e.g., surface material and types of equipment) should also be examined in future research. Moreover, we elected to dichotomize playgrounds into high (>8) and low (≤8) quality to examine how proximity to superior playgrounds is associated with youth weight status; however, future studies may wish to analyze the contribution of each additional quality element to reducing the odds of overweight and obesity as might be elucidated using a more continuous measure. Likewise, it would be interesting to consider the moderating influence of demographic characteristics (e.g., individual or neighborhood SES or race/ethnicity) on the association between playground quality and weight status. Furthermore, a network buffer of ½ mile was used to represent walking or bike riding distance from a child's home. Other distances may be relevant in associating park and playground access with youth obesity and other factors within the neighborhood (e.g., crime, traffic, and sidewalks), which may also impact youths' use of available playgrounds. Finally, in examining playground quality, average playground scores were assigned to children with more than one playground within their buffer; however, these mean scores may obfuscate large variations in quality for nearby amenities.

Conclusion

In summary, this was one of the first studies to examine both disparities in detailed elements of playground quality and the relationship between playground access and quality and youth weight status. Playgrounds are key features for youth PA participation, and more research is warranted to better understand issues related to their accessibility and design for different user groups. For example, qualitative research could explore the relative importance of the factors studied herein to potentially adjust the weighting of the various quality attributes we examined. In addition, longitudinal studies should examine how modifications to playground quality characteristics impact the attractiveness of play spaces to both youth and parents and the ultimate impact on energy expenditure and obesity among users. For example, Colabianchi et al. reported that renovated playgrounds (which assumedly improved their perceived quality) were associated with more usage and higher intensity activity,¹⁶ while Roemmich et al. reported that the temporary removal of benches at playgrounds resulted in greater activity levels among supervising parents and other spectators.³⁹ While such microscale park interventions will vary in scope and expense, greater attention should be devoted to understanding the cost-effectiveness of playground quality modifications relative to gains and losses, respectively, in youth and family PA and obesity levels. Finally, the quality attributes we examined in this study can be combined with other limited research documenting how specific activity spaces within playgrounds (e.g., swings) are associated with youth PA levels.¹⁹ Having a comprehensive picture of playground access and quality and their relationship with weight status among youth from diverse backgrounds can lead to healthier communities for all children.

Footnotes

Acknowledgments

Piedmont Health Foundation, The University of South Carolina Office of the Vice President for Research, and University of South Carolina Office of the Provost.

Author Disclosure Statement

No competing financial interests exist.

References

Ogden

, Carroll

, Fryar

, Flegal

. Prevalence of obesity among adults and youth: United States, 2011–2014. NCHS Data Brief, 2015;(219):1–8.

Dehghan

, Akhtar-Danesh

, Merchant

. Childhood obesity, prevalence and prevention. Nutr J, 2005; 4:24.

Troiano

, Berrigan

, Dodd

, et al. Physical activity in the United States measured by accelerometer. Med Sci Sports Exerc, 2008; 40:181–188.

Sallis

, Floyd

, Rodríguez

, Saelens

. The role of built environments in physical activity, obesity, and CVD. Circulation, 2012; 125:729–737.

Ding

, Sallis

, Kerr

, et al. Neighborhood environment and physical activity among youth: A review. Am J Prev Med, 2011; 41:442–455.

Dunton

, Almanza

, Jerrett

, et al. Neighborhood park use by children: Use of accelerometry and global positioning systems. Am J Prev Med, 2014; 46:136–142.

Potwarka

, Kaczynski

, Flack

. Places to play: Association of park space and facilities with healthy weight status among children. J Community Health, 2008; 33:344–350.

Qazi

. Childhood obesity and parks and playgrounds: A review of issues of equality, gender and social support. J Res Med Sci, 2011; 16:553–558.

Bedimo-Rung

, Mowen

, Cohen

. The significance of parks to physical activity and public health: A conceptual model. Am J Prev Med, 2005; 28(2 Suppl 2):159–168.

10.

Alexander

, Huber

LRB

, Piper

, Tanner

. The association between recreational parks, facilities and childhood obesity: A cross-sectional study of the 2007 National Survey of Children's Health. J Epidemiol Community Health, 2013; 67:427–431.

11.

Wolch

, Jerrett

, Reynolds

, et al. Childhood obesity and proximity to urban parks and recreational resources: A longitudinal cohort study. Health Place, 2011; 17:207–214.

12.

Timperio

, Giles-Corti

, Crawford

, et al. Features of public open spaces and physical activity among children: Findings from the CLAN study. Prev Med, 2008; 47:514–518.

13.

Besenyi

, Kaczynski

, Wilhelm Stanis

, Vaughan

. Demographic variations in observed energy expenditure across park activity areas. Prev Med, 2013; 56:79–81.

14.

Black

, Menzel

, Bungum

. The relationship among playground areas and physical activity levels in children. J Pediatr Health Care, 2015; 29:156–168.

15.

Quigg

, Gray

, Reeder

, et al. Using accelerometers and GPS units to identify the proportion of daily physical activity located in parks with playgrounds in New Zealand children. Prev Med, 2010; 50:235–240.

16.

Colabianchi

, Kinsella

, Coulton

, Moore

. Utilization and physical activity levels at renovated and unrenovated school playgrounds. Prev Med, 2009; 48:140–143.

17.

Cradock

, Kawachi

, Colditz

, et al. Playground safety and access in Boston neighborhoods. Am J Prev Med, 2005; 28:357–363.

18.

Arroyo-Johnson

, Woodward

, Milam

, et al. Still separate, still unequal: Social determinants of playground safety and proximity disparities in St. Louis. J Urban Health, 2016; 93:627–638.

19.

Colabianchi

, Maslow

, Swayampakala

. Features and amenities of school playgrounds: A direct observation study of utilization and physical activity levels outside of school time. Int J Behav Nutr Phys Act, 2011; 8:32.

20.

Hannon

, Cradock

, Gortmaker

, et al. Play across Boston: A community initiative to reduce disparities in access to after-school physical activity programs for inner-city youths. Prev Chronic Dis, 2006; 3:A100.

21.

Gordon-Larsen

, Nelson

, Page

, Popkin

. Inequality in the built environment underlies key health disparities in physical activity and obesity. Pediatrics, 2006; 117:417–424.

22.

Hoefer

, McKenzie

, Sallis

, et al. Parental provision of transportation for adolescent physical activity. Am J Prev Med, 2001; 21:48–51.

23.

Besenyi

, Kaczynski

, Wilhelm Stanis

, et al. Sex differences in the relationship between park proximity and features and youth physical activity. Child Youth Environ, 2016; 26:56–77.

24.

United States Census Bureau.. State and quick county facts. 2012. Available at www.census.gov/quickfacts/table/PST045215/00 (last accessed June 2, 2013).

25.

About Child & Teen BMI | Healthy Weight | DNPAO | CDC. Available at www.cdc.gov/healthyweight/assessing/bmi/childrens_bmi/about_childrens_bmi.html Last accessed March 29, 2016.

26.

Cohen

, Ashwood

, Scott

, et al. Public parks and physical activity among adolescent girls. Pediatrics, 2006; 118:e1381–e1389.

27.

Loukaitou-Sideris

, Sideris

. What brings children to the park? Analysis and measurement of the variables affecting children's use of parks. J Am Plann Assoc, 2009; 76:89–107.

28.

Kaczynski

, Stanis

SAW

, Besenyi

. Development and testing of a community stakeholder park audit tool. Am J Prev Med, 2012; 42:242–249.

29.

Community Park Audit Tool (CPAT) | Active Living Research. http://activelivingresearch.org/community-park-audit-tool-cpat Last accessed June 30, 2016.

30.

Hughey

, Walsemann

, Child

, et al. Using an environmental justice approach to examine the relationships between park availability and quality indicators, neighborhood disadvantage, and racial/ethnic composition. Landsc Urban Plan, 2016; 148:159–169.

31.

Abercrombie

, Sallis

, Conway

, et al. Income and racial disparities in access to public parks and private recreation facilities. Am J Prev Med, 2008; 34:9–15.

32.

Kamel

, Ford

, Kaczynski

. Disparities in park availability, features, and characteristics by social determinants of health within a U.S.-Mexico border urban area. Prev Med, 2014; 69 Suppl 1:S111–S113.

33.

Edwards

, Hooper

, Knuiman

, et al. Associations between park features and adolescent park use for physical activity. Int J Behav Nutr Phys Act, 2015; 12:21.

34.

Kaczynski

, Stanis

SAW

, Besenyi

, Child

. Differences in Youth and Adult Physical Activity in Park Settings by Sex and Race/Ethnicity. Prev Chronic Dis, 2013; 10:E42.

35.

Vaughan

, Kaczynski

, Wilhelm Stanis

, et al. Exploring the distribution of park availability, features, and quality across Kansas City, Missouri by income and race/ethnicity: An environmental justice investigation. Ann Behav Med, 2013; 45 Suppl 1:S28–S38.

36.

Thornton

, Conway

, Cain

, et al. Disparities in pedestrian streetscape environments by income and race/ethnicity. SSM Popul Health, 2016; 2:206–216.

37.

Kaur

, Lamb

, Ogden

. The association between food insecurity and obesity in children—The National Health and Nutrition Examination Survey. J Acad Nutr Diet, 2015; 115:751–758.

38.

Tucker

, Gilliland

, Irwin

. Splashpads, swings, and shade: Parents' preferences for neighbourhood parks. Can J Public Health, 2007; 98:198–202.

39.

Roemmich

, Beeler

, Johnson

. A microenvironment approach to reducing sedentary time and increasing physical activity of children and adults at a playground. Prev Med, 2014; 62:108–112.