Territorial inequalities in the food environment of Brazilian metropolitan regions.

OBJECTIVE: To describe the retail food environment of Brazilian metropolitan regions using intramunicipal socioeconomic indicators, applying the Locais-NOVA classification. METHODS: Data from the 2023 Annual Social Information Report were used, with Human Development Units (HDUs) as the unit of analysis. Food establishments were geocoded and assigned to HDUs. The Locais-NOVA methodology was applied, which classifies food acquisition outlets according to the degree of food processing and proposes five food environment indicators. The percentage of each indicator was calculated for HDUs and stratified into tertiles of the Municipal Human Development Index (MHDI), considering the distribution of HDUs, Brazilian macroregions, and capitals versus non-capitals. A food environment healthiness indicator was created to measure the competition between healthy and unhealthy outlets and was used for spatial description through maps. RESULTS: A total of 8,977 HDUs across 24 metropolitan regions were assessed, with a mean MHDI of 0.76 and an average of 42.8 food establishments per HDU. Most food outlets were classified as sources of ultra-processed foods (64.9%), with a higher proportion in the first MHDI tertile (67.5%) compared with the third tertile (63.1%). A lower presence of outlets sourcing fresh and minimally processed foods was observed in the first MHDI tertile compared with the third (25.8 vs. 32.3%). The Southeast showed the highest percentage of outlets sourcing processed foods (17.3%). Non-capital concentrated a higher proportion of outlets sourcing processed foods compared with capitals. CONCLUSION: A high prevalence of outlets sourcing ultra-processed foods was observed, particularly in territories with worse socioeconomic conditions.