Comunicaciones e-póster


Vanessa Cardozo Mendes Elias1, Mônica de Caldas Rosa dos Anjos1, Sandra Patricia Crispim1.

1Universidade Federal Do Paraná, Curitiba, Brazil.

Background and objective. To assess the consumption of ultra-processed foods, the NOVA classification was created. The aim of the present study was to verify the presence of information on food processing in food surveys and their contribution to the NOVA classification.

Methods. For this purpose, data were collected from dietary surveys previously collected in four Brazilian studies that contemplate different methods of data collection. The foods were grouped according to a NOVA classification, and also in a group called ‘uncertainty’. The daily energy intake of each group was estimated, as well as the energy contribution. Three different scenarios were created according to the degree of uncertainty. preliminary, most probable, and counter-hypothesis, which assumes the lowest possible degree of processing for foods that were described in a generic way and remained uncertain in the classification.

Results. The descriptors that indicate some kind of food processing were present in 29.2; 3.3; 4.2 and 24.4% of the foods mentioned in VALIDA, POF, QFA-SUL, and EMDI respectively. The average contribution from the uncertainty group in VALIDA, POF, QFA-SUL, and EMDI was 38.6; 35.9; 34.1, and 26.7% of the total consumed, respectively. Considering the three scenarios, it appears that the evolution of the uncertainties from the QFA-SUL study was not as expressive as in the other studies. In the EMDI study, even in cases of uncertainty, more elements were provided to assume what was most likely for each case, which led to a large reduction in uncertainty in the most probable scenario. When descriptors were combined, the food processing information became more precise.

Conclusions. It is important that foods with an ultra-processed industrial equivalent are carefully collected, as it is the group with larger doubts. Different scenarios in the evaluation of food consumption can influence the results, under or overestimating the participation of specific groups of NOVA in the total energy intake.