Leveraging cell phone location data to measure interactions with the food environment and associated health outcomes

Project Summary Unhealthy diet is a key behavioral risk factor for several chronic diseases, particularly type-2 diabetes, cardiovascular disease, and some cancers. Long-term measures of diet are typically self-reported and subject to known sources of error and bias. A recognized population-level determinant of diet is access to and use of

food retailing, particularly consuming food away from home (FAFH) from quick-service or “fast food” outlets. Currently, there are few methods to objectively measure long-term patterns of FAFH in health research or to understand how built environments shape these behaviors. However, the pervasive use of smartphones provides

an emerging opportunity to address these questions. We will use Google Location History Timeline (GTL) data, which includes high-resolution geographic location data collected passively on smartphones with Google Apps, to develop new measures of long-term patterns of food retailer utilization, especially use of FAFH outlets. We

will use data from 500 adults from the Washington State Twin Registry who are contributing their GTL data for another NIH-funded study on built environments and active travel (R21 ES031226-01A1). These data include daily time-activity patterns from 2012 to present. We will evaluate these measures against objective measures

of contemporaneous food outlet utilization and detailed dietary intake measures in a subset of twins who have previously completed additional surveys, including food frequency questionnaires and travel diaries. Finally, we will measure how metrics of FAFH use relate to health outcomes, particularly weight status, weight change, and

self-rated overall health, both cross-sectionally and longitudinally. New measures of food access behaviors associated with diet and health outcomes can be used in epidemiological research and advance the field of built environment influences on human health.