Characterizing household food shopping behavior using novel receipt‐based geospatial measures

Abstract

BackgroundIn efforts to better understand how geographic food access measures influences dietary behaviors, researchers are developing second‐generation methodologies to explore and track how individuals behave within their food environment.MethodsUsing receipt‐based data collected over four weeks in 90 households in Minneapolis/St. Paul, MN and geographic information systems (GIS), we created two geospatial measures to characterize households' shopping spaces via minimum bounding geometry. Thus, a spatial ‘footprint’ representing the geographic areas households purchased food at supermarkets and restaurants was determined. Statistical correlation with traditional measures of shopping behavior (frequency and distance traveled) were assessed along with associations with food purchasing (in dollars) and dietary intake of fruit, vegetables, and sugar‐sweetened beverages.ResultsOn average, households' shopping space for supermarkets and restaurants were 15.9 and 41.8 square miles, respectively. Shopping space was significantly correlated with both frequency of shopping and distance traveled for supermarkets (r=0.59 and r=0.54) and restaurants (r=0.74 and r=0.45). Neither shopping space for supermarkets or restaurants were significantly associated with purchases of fruits and vegetables, however, restaurant shopping space was significantly associated with expenditures for sugar‐sweetened beverages (β=0.06, p=0.004) and vegetable intake (cup servings) during the past 7 days (β=‐0.01, p=0.015).ConclusionsFindings suggest that household shopping space may capture shopping and purchasing behaviors of interest to researchers examining the relationship between the food environment and dietary intake.