Food outlet availability differs according to area socioeconomic status in Denmark

BACKGROUND: Disparities in access to specific foods may contribute to inequalities in diet-related diseases seen at a global and National level.

METHODS: Based on aggregated population data on income, education, and employment, area-level socioeconomic differences in food outlet availability were analyzed for 53,368 study participants residing across parishes in the Capital Region of Denmark. Validated data on fast-food outlets, convenience stores, supermarkets, and restaurants were used. Information on individual characteristics, home address and corresponding parishes were linked to the participants through the Danish Civil Registration System. Three multilevel hurdle models were applied for each food outlet type to analyze food outlet density (count/km²) within an 800-meter network buffer around participants' homes across four levels of parish socioeconomic status (SES). Model 1 provided a basic examination of the association between density and area SES. Model 2 adjusted for individual characteristics while Model 3 further included urbanity at the area level. The structure of the chosen hurdle models included was Part (1) a logistic multilevel regression modelling the probability of food outlet presence by using the entire dataset and, Part (2) a standard linear multilevel regression modelling the 10 base logarithmic transformation of only positive food outlet densities with a lognormal distribution.

RESULTS: No statistically significant spatial patterning of food outlets across area SES was found in Model 1 and 2, however positive and strong significant odds were seen in part 1 of Model 3 for supermarkets, convenience stores and fast-food outlets. Thus, residents in more disadvantaged SES areas had higher odds of having a supermarket, convenience store, or fast-food outlet near their homes compared to those living in the most advantaged areas. No differences were seen in the density across area SES.

CONCLUSION: Area SES influenced the presence of supermarkets, convenience stores, and fast-food outlets, but not the density of these establishments.