LAUSR

OBJECTIVE Investigate whether the combined consumption of fresh/minimally processed and ultra-processed food is associated with food insecurity during the COVID-19 pandemic. DESIGN Cross-sectional observational study was derived from a survey using a population-based search of a complex sample. Food insecurity was assessed using the validated Brazilian Food Insecurity Scale. Food consumption was assessed using a qualitative food frequency questionnaire and the NOVA classification for fresh/minimally processed food and ultra-processed food. A scoring system was used to evaluate combined food consumption according to the extent and purpose of processing, considering the weekly consumption of the two groups (according to the NOVA classification). Higher punctuation reflects worse diet quality (higher consumption of ultra-processed foods and lower consumption of fresh/minimally processed foods). A theoretical causality model was constructed using a directed acyclic graph, and multivariate analysis was performed using Poisson regression to test the association between food insecurity and food consumption. SETTING Ouro Preto and Mariana, Brazil, between October and December 2020. PARTICIPANTS An epidemiological household survey was conducted with 1,753 individuals selected through a stratified and clustered sampling design in three stages. RESULTS Those with food consumption scores in the fourth quartile had a 60% higher prevalence ratio (PR) for food insecurity (RP: 1.60 and 95%IC: 1.06 - 2.40). Also, the increased consumption of fresh/minimally processed foods and low consumption of ultra-processed foods presented a 45% lower prevalence ratio of food insecurity (RP: 0.55 and 95%IC: 0.40 - 0.80). CONCLUSION These results indicate an inverse association between food insecurity and diet quality.