A Study of Neighborhood Park Use and Children’s Total Physical Activity Using GPS and Accelerometer Data

Background Research has suggested that green and natural spaces within the built environment are associated with increased physical activity in children. in particular, parks have been identified as important features in the community for promoting active living. Presence of a neighborhood park may encourage active living simply by drawing residents outdoors. It also promotes activity by providing a space wherein children can play. It is not well known, however, to what extent neighborhood park use may contribute to children‘s total physical activity behavior.

Objectives To examine whether presence of a neighborhood park or park use is associated with children‘s total physical activity behavior.

Methods The study sample included 135 children drawn from a larger cohort study in Southern California, Healthy PLACES. Children were of ages 8-14, 50% female, and represented a wide range of socioeconomic backgrounds and multiple ethnic groups (48% Hispanic). Study participants were asked to wear a GlobalSat BT335 portable global positioning system (GPS) and an Actigraph accelerometer for seven days to measure date, time, location, and physical activity. Both devices were pre-programmed to take measures every 30 seconds. Since the focus of this study was on the relationship between neighborhood parks and leisure time activity, GPS-activity data collected during school hours (9am – 2pm weekdays during the school season) were excluded. A geographical information system, ArcGIS 9.3 (ESRI, Redlands, CA), was used to estimate proximity of GPS-activity points to park spaces within the neighborhood (parks within 500m of home). Children who spent more than 5 minutes within 10 meters of the park space where more than 1 of those minutes were spent within the park space were classified as having used their neighborhood park (park use). Accelerometer activity data were further classified as moderate or vigorous (MVPA) by using age-specific cut-points for corresponding predicted metabolic equivalents. A child was determined to have used their neighborhood park for MVPA if they engaged in more than 5 minutes of MVPA within the park space (park-MVPA). A log-linear regression model was then applied to assess whether binary predictors, presence of a neighborhood park, park use, or park-MVPA, were associated with the average daily amount of neighborhood-MVPA (MVPA occurring within 500m of home), as well as the average daily amount of MVPA overall, while controlling for age, gender, race, and household income.

Results Home addresses for 135 children were geocoded in ArcGIS and neighborhood park spaces (within 500m of home) were identified. Approximately half of the children had a park in their neighborhood. Regression results suggest that on average children with a neighborhood park experienced almost twice as much daily neighborhood-MVPA (parameter estimate=1.95, 95% CI=1.08, 3.51) while controlling for age, gender, race, and household income. in contrast, neighborhood park presence was not significantly associated with the amount of daily MVPA overall (p-value > 0.05). Further analysis on the subset of children with a park (n=73) found that GPS-determined park use was associated with increased daily neighborhood-MVPA (p-value < 0.001) as well as increased daily MVPA overall (p = 0.02). Specifically, children who used their available park space had almost four times as much daily neighborhood-MVPA (parameter estimate = 3.98, 95% CI = 2.02, 7.82), and almost twice as much daily MVPA overall (parameter estimate = 1.80, 95% CI = 1.11, 2.92), as compared to children who did not use their neighborhood park space. an additional analysis on this subset of children who had a neighborhood park suggests that those who engaged in > 5 minutes of MVPA within the park space (park-MVPA) had on average almost five times the amount of neighborhood-MVPA per day (parameter estimate = 4.80, 95% CI = 2.26, 10.20), and twice as much daily MVPA overall (parameter estimate = 2.02, 95% CI =1.18, 3.46), as compared to children who had a park but did not engage in MVPA there either because they did not visit their local park or because they were not physically active in the park.

Conclusions Study results suggest that neighborhood park presence and use are positively associated with children‘s physical activity behavior. Children who used their neighborhood park space experienced four times more daily MVPA in their neighborhood, and almost twice as much daily MVPA overall, compared to children who had a park but did not spend time there. These findings highlight the contribution park space activity may have to children‘s physical activity behavior and provide strong support for the integration of park spaces into community design to promote active living. Study results also highlight the need to encourage residents to use their available park spaces in the neighborhood. Future analyses will include the complete Healthy PLACES cohort of 609 children, and will further examine these activity patterns at parks located both inside and outside of neighborhoods.

Support/Funding Source SUPPORT: UC Berkeley Chancellor‘s Fellowship, the CDC-ASPH Environmental Health Scholarship, and the National Cancer Institute grant #R01-CA-123243 (Pentz, PI).