Use of GIS to Measure Food Environment and Its Relationship with Obesity in School-aged Children in Mexico (P04-142-19)

Abstract

ObjectivesTo determine the association between food environment and obesity in school-aged children using Geographic Information Database (GIS). MethodsA total of 218 school-age children (8.4 ± 1.5 y) from a low-income setting in Mexico participated in this cross-sectional study. Weight and height were measured in all children to calculate BMI- for-age z-score (BMIz). Body composition (total body fat and abdominal fat) was measured using DXA. Geolocation of both convenience stores (CS) and participants’ households was collected and introduced into a GIS database. The shelf-space of processed foods and unprocessed foods available at each CS was measured. The distance to the closest CS, the number of CS, and the shelf-space of processed foods of CS within a 250 m radius from each participant’s household was calculated using GIS. Multivariable model assessing the association between obesity and food environment were performed controlling for confounders, using SPSS v23.0. ResultsMore than half of the children (55%) were girls, and 52% of the children had elevated body fat % (>30% for girls and >25% for boys). A total of 58 food venues were found in this community; from these venues, 40 were CS. 91% of the children had access to a CS within a 250 m radius. In average, 48% of the shelf-meters of the CS were of processed food. A positive association between BMIz (β = 0.028; IC 95%: 0.005,0.047; P < 0.05), total body fat % (β = 0.145, IC 95%: 0.048, 0.241; P < 0.05) and abdominal fat % (β = 0.206, IC 95%: 0.069,0.343; P < 0.05) was found with the distance to the nearest CS within the 250 radius. However, a negative association was found between BMIz (β = −0.002; IC 95%: −0.004, −0.001; P < 0.05), total body fat % (β = −0.009, IC 95%: −0.017, −0.001; P < 0.05) and abdominal fat % (β = −0.012, IC 95%: −0.023, −0.001; P < 0.05) with the proximity of CS. Finally, BMIz was positively associated with the meters of processed foods in the CS within 250 m radius (β = 0.002; IC 95%: 0.001,0.003; P < 0.05). ConclusionsA high density and proximity of convenience stores with a high availability of processed food were associated with obesity in children of this community. GIS methods may provide a new approach to assess food environment and its association with obesity risk factors in different settings and populations. Funding SourcesCONACYT.