The last several years have witnessed growing interest in the potential contribution of the ?built environment? to the epidemic increase in obesity rates in the United States. This includes both the large structural aspects of the neighborhood and physical characteristics of homes and schools at the very micro level (the interiors and associated grounds of these buildings). The structural aspects of the neighborhood that surround the home and school includes urban design features such as streets and highways, land use mix, street connectivity, public transportation infrastructure (i.e., subway stations), park land, and recreational infrastructure, all of which could affect both physical activity and the accessibility of food and other resources. The physical characteristics of the home or school?including building size, the presence of elevators, stairs, spaces for physical activity, and related factors?could determine opportunities for physical activity and amount of activity needed to navigate the space. Thus far not established is how each of these factors contributes to childhood BMI in a distinct and interconnected way and/or how changes in such factors could influence obesity rates in a causal manner. This gap in knowledge is particularly troubling given the high cost of changing the built environment as well as the longevity of investments once made. Our research promises to provide insight critical to policymakers planning urban infrastructure investment. Particularly important to consider are disparities, and whether differences in the built environment could explain differences in obesity rates in key demographic groups, including income and race/ethnicity. We propose to combine New York City Department of Education FITNESSGRAM data, which includes BMI, school and residential locations, for all New York City public school children from 2005 onward, with detailed data on the built environment surrounding each child?s home and school as well as within their home and school. Then, we will use a variety of enhanced methodological techniques to estimate a less-biased, more causal relationship between the built environment and obesity. The methodological enhancements we utilize over past work are: ? A large, detailed dataset, including data on over 1 million NYC public school children. ? Examining longitudinally the same children over time, using both child, school and census tract level fixed effects, among other methods, to determine casualty. ? Examining fine-grained differences in the built environment. For example, the relative influence of living within 500 feet of a park or other built environment resources ? Data on both the home and school built environment, both inside the building and in the neighborhood, modeled together. With these enhancements, we will be able to provide improved estimates on the influence of the built environment on child BMI, substantially answering key health and policy-related questions.
