Drexel is providing the leadership of this project and is responsible for coordination of file sharing. Data quality and documentation have been ensured by a team, which spans Drexel University Urban Health Collaborative, Columbia University Built Environment and Health team, the University of Washington, University of Alabama at Birmingham, Environment Policy Lab at University of Michigan.

Assembled datasets include

• Census tracts and ZCTAs throughout the contiguous US (using 2010 boundaries).

  • 72,246 census tracts throughout the continental US

  • 32,170 ZIP code tabulation areas (ZCTA) throughout the continental US

• 1km and 5km circular buffers around resident addresses (CHS and REGARDS). Residential addresses were supplemented using Lexis Nexis and geocoded (Lexis Nexis personal profile data also cover voting, property ownership, etc).

  • 13,790 1-km and 13,790 5-km radial buffers centered on home addresses of participants from the Cardiovascular Health Study (CHS)

  • 41,515 1-km and 41,515 5-km radial buffers centered on the home addresses of participants from the REasons for Geographic And Racial Differences in Stroke (REGARDS)

For NETS Data

We created a protocol for classifying potentially health-relevant establishments from commercially licensed established data. Our categorization of annual retail location data from the National Establishment Time Series dataset (1990-2014) used 8-digit Standard Industrial Classification codes, informed by prior literature. Systematic spot checks were conducted across 30 locations of varying urbanicity in all 10 US census regions (to inform the inclusion or exclusion of 81 SIC codes in the health-relevant categories establishments). Our protocol was presented at the 2020 Society for Prevention Research, and a manuscript describing our protocol is currently being revised based on reviewer comments.

Based on the rigor and transparency of our methods, we have been invited to collaborate with several federally funded research projects including work with

1) the Census Headquarters staff to the NIH funded Mortality Disparities in American Communities data (4 approved manuscript proposals with writing leads at Drexel University using this data resource will be shared in a panel presentation at the 2020 IAPHS conference) and

2) others who have grant funding to characterize the geographic contexts surrounding REGARDS or CHS participants [Natalie Colabianchi (Environment and Policy Lab, University of Michigan), Michelle Carlson (Brain Health Lab, Johns Hopkins University)]

3) with the CDC-funded Diabetes LEAD Network assess the food environment, physical activity, and medical facilities in residential neighborhoods for

As research on healthy aging and social determinants of health emphasize neighborhood context, our advanced novel method we established for preparing longitudinal retail and commercial data will be shared with others (within the terms of confidentiality and data licensing agreements). This work promotes efficiency in working with complex data, facilitating previously unattainable avenues of longitudinal and national research and policy.