Lovasi is an associate professor and co-director of the Urban Health Collaborative in the Dornsife School of Public Health.
Where you live should not determine your lifespan, but it often does.
Health disparities in neighborhoods result from access to transportation, limited supplies of nutritious food, poor access to primary care and unstable housing, among other factors. To address these issues, it’s critical to assess all the factors impacting the health of a community — and that’s where big data comes in.
At the Dornsife School of Public Health, Associate Professor of Urban Health and co-director of the Urban Health Collaborative Gina Lovasi is culling information from some of the nation’s largest databases on neighborhood resources — from stores, to restaurants, parks and pools.
Lovasi’s team has done so much to clean up the retail environment data that other public health researchers are asking to use it.
Lovasi’s team is mapping business of all types to create nationally linked datasets under an effort called the Retail Environment and Cardiovascular Disease study. They licensed and geocoded comprehensive business data from across the continental United States for each year in the period from 1990 to 2014, and created more than 100 health-relevant retail categories. Their goal is to use neighborhood data alongside health outcomes to inform zoning initiatives, policies and public health programs that increase healthy food availability, walkability and proximity to medical facilities.
They’ve done so much to clean up the accuracy of the data that other groups are asking for the resource. For instance, the CDC-funded Diabetes LEAD Network, with partners from Drexel, Geisinger-Johns Hopkins University, New York University and University of Alabama, will use the retail environment data to understand geographic patterns of diabetes risk and related health complications. In addition, the National Institutes of Health funded an extension of the project to investigate new aims on independent aging and cognitive function.
“One of the major limitations of earlier work [of this type] was relying on a snapshot in time,” Lovasi says. “We’re [now] able to look over a long period for changes. This helps us make sure we’re measuring the environment at the right time. We are also looking at trajectories of change and whether those trajectories affect health. That’s why we want to go back to 1990. It takes decades to accumulate enough cases to do this work.”