Abstract
OBJECTIVES: To adapt and validate a tool to measure the level of food knowledge based on the Nova classification for the Brazilian context. METHODS: A tool developed by Canadian researchers was adapted for Brazil. In this tool, respondents assign healthiness scores to 12 images of foods with different levels of industrial processing according to the Nova classification - unprocessed and minimally processed, processed and ultra-processed. Total score is computed by comparing scores assigned to foods from different groups, and range from 0 to 8. The Brazilian version, named Nova-Conhecimento, was evaluated by experts and submitted to pre-tests with potential users. Discriminant validity was assessed by comparing scores of undergraduate students of nutrition and undergraduate students in education-related fields. Convergent validity was assessed by testing the association between the knowledge score and the consumption of ultra-processed foods in a subsample of the NutriNet Brazil cohort (n=1,245). RESULTS: Nutrition students had higher scores than education students (6.7 vs. 5.3; p-value<0.001). Each point in the knowledge score was associated with a reduction of 1.03 percentage points in the contribution of ultra-processed foods to the diet (p-value<0.001). CONCLUSION: The Nova-Conhecimento tool demonstrated validity and can contribute to food and nutritional surveillance activities based on the the Dietary Guidelines for the Brazilian Population.