Using geospatial technologies to explore activity-based food environments: Considering the duration of GPS tracking

Researchers often examine the types of food stores and restaurants located within neighborhoods to describe the food environments of individuals. Most Americans, however, travel daily outside their neighborhoods. This paper describes a study conducted in Lexington, KY that collected global positioning system (GPS) data to investigate food environments within participants' activity spaces, or the spaces in which they conduct daily activities. Furthermore, it explores the influence of GPS tracking duration on these activity-based food environment assessments.

One-hundred twenty-one residents of a census tract carried a GPS data logger for three weekdays and answered a survey regarding diet and food shopping. Thirteen later participated in GPS tracking for a full week. Food locations were obtained from the county health department to characterize the food environment within each participant's activity space, defined as a half-mile buffer around their GPS track. Statistical analysis included t-tests, ANOVA, and pairwise correlations to compare food environment measures and participant characteristics.

The main findings of this study indicated several statistically significant relationships among participants' diets, food purchases, and activity-based food environments. GPS tracking duration, however, significantly influenced activity space size and measures of the food environment. Seven-day activity spaces were larger (10.14 vs. 20.45 sq mi, p<0.0001), and contained lower densities of full-service (8.69 vs. 6.42 per sq mi, p=0.01) and limited-service restaurants (12.43 vs. 9.25, p=0.01), small groceries (1.59 vs. 0.88, p=0.01), and convenience stores (2.31 vs. 1.81, p=0.01). Proportional values were more similar. Future studies should carefully consider the effects of GPS tracking duration.