Save the Date!
ActiGraph Digital Data Summit 2021November 4 - 5 | Register for Event Updates Now
GIS Environmental Determinants of Objectively Measured Physical Activity and Intervention Uptake in Children
- Presented on 03/01/2011
Background Although physical activity fosters healthy development among children (such as reducing obesity and type 2 diabetes risk as well as promoting positive mental well-being), less than half of children meet national physical activity guidelines. The accumulating research indicates that neighborhood environmental characteristics (e.g. access to walking destinations and community design) may facilitate or impede physical activity among children. However, the vast majority of studies in this area exclusively rely on self-reported physical activity measures and do not evaluate the scale or zoning sensitivity of the neighborhood definition used. Additionally, little is known empirically about environmental influences on intervention uptake.
Objectives The primary aim of this investigation was to examine gender-stratified relationships between neighborhood environment characteristics and physical activity by different neighborhood definitions (i.e. 200-meter, 400-meter and 800-meter street network and circular buffers). A secondary aim was to examine if neighborhood environment factors impact the uptake of an after-school obesity prevention intervention on physical activity level among children.
Methods Data come from the YMCA-Harvard After School Food and Fitness Project, a multi-site quasi-experimental intervention targeting children ages 5-12 and their families focused on changing after school environments to promote physical activity and healthy eating. The analysis to address the first study aim include baseline data collected from children who had georeferenced home addresses and at least two days of accelerometer data (N=407). For our second aim, relationships were evaluated for children in the intervention group who met the criteria for the first study aim and had complete baseline and follow-up accelerometer data with at least two days for both time points (N=201). We built a spatial database using geographic information systems (GIS) technology with ArcGIS 9.3, creating the following environmental variables: parks per sq km; walking destinations for retail, services, and cultural/educational activities per sq km; walking destination mix; residential density per sq km; intersections per sq km; median pedestrian route directness; and cul de sac presence. GIS data layers came from Info USA. Minutes per day of moderate-to-vigorous physical activity is the primary outcome and overall physical activity (mean accelerometer counts per minute) is the secondary outcome. We used objective physical activity data collected with ActiGraph model 7164 accelerometers. Other variables included multiple individual and neighborhood characteristics. We first conducted exploratory spatial data analysis, including visualization and cluster detection. Visualization was conducted in ArcMap 9.3 to map key study variables and to visually examine patterns across space. Then, global cluster detection was conducted with a Global Moran’s I to evaluate whether there is spatial autocorrelation for the outcomes. Next, we fit ordinary least squares (OLS) regression and spatial autoregressive models as appropriate. For example, if the residuals of the OLS were not significant for spatial auto-correlation, there was no indication that a spatial model is needed. If evidence for a spatial effect was found, spatial error regression models, estimated via maximum likelihood, were fit. The spatial weight matrix, which provides the structure of assumed spatial relationships, was specified as k nearest neighbors (n=4). The models were adjusted for potential confounders and clustering of children within after school sites. To examine our second aim, if the after-school intervention uptake was influenced by the neighborhood environments of the participants, we limited the sample to only those in the intervention group and re-estimated regressions equations predicting change in physical activity levels with the neighborhood environment characteristics assessed in aim 1 found to be significant predictors, adjusting for baseline covariates as in aim 1 as well as baseline physical activity. In follow-up analysis, the regression coefficient represents intervention uptake, operationalized as the difference between of physical activity from baseline to follow-up.
Results Preliminary analysis of one of the four after school sites indicate that the Global Moran’s I for average accelerometer counts was low and not significant. We anticipate that access to walking destinations as well as community design will be positively associated with both types of physical activity, consistent with current literature. We also anticipate that these associations will vary by gender and vary depending on the definition of the neighborhood used. Lastly, we anticipate that the neighborhood characteristics will influence intervention uptake. All analysis will be completed by fall 2010.
Conclusions This study may lend support to the burgeoning research indicating that neighborhood characteristics are implicated in physical activity among children, which has substantive neighborhood-level policy relevance, especially because the neighborhood characteristics we examined are: most consistently and strongly associated with physical activity among children, and amenable to policy change. This study also addresses methodological issues and fills several gaps that exist in the extant research literature, including related to the sensitivity of neighborhood effects by explicitly addressing the modifiable areal unit problem – demonstrating to health researchers that critical thought needs to be given when defining neighborhoods.
Support This project is funded by an Active Living Research Dissertation Award from the Robert Wood Johnson Foundation (Grant # 67129).