The aim of this study is to have an exploratory insight on how a sample of Brazilian adults classify food, attempting to identify the main factors involved in this process, and to compare these classifications to the NOVA food classification of the 2014 Brazilian dietary guidelines. An exploratory and qualitative approach was conducted with a selected sample of teachers, administrative technicians, and students (N = 24) from the Federal University of Grande Dourados, Brazil. First, using the pile sort method, participants were asked to freely classify 24 pictures of food (sourced from examples of the four food groups specified in NOVA) into food groups meaningful to them. Next, in semi-structured interviews, participants were asked to describe the food groups they created. The food groups created by participants were analyzed using non-metric multidimensional scaling followed by hierarchical cluster analysis, and the interviews were analyzed using content analysis. Participants had a mean age of 30 (±9.4) years. A total of 128 food groups were created by 24 participants (an average of five food groups per person); and a total of 55 non-mutually exclusive groups names were used by them to describe these food groups. Sixteen themes emerged from the content analysis. The most recurrent themes were food groups, nutrients, foods I consume, foods I do not consume, and food processing. Contrasting themes such as real food and junk foods, meals and ready-made foods, healthy foods and unhealthy foods were also noted. Six clusters emerged from the cluster analysis, each related to one or more themes. Overall, a striking similarity was observed between the ways the individuals classified food and the NOVA food classification.
Keywords: Dietary guidelines; Food classification; NOVA; Pile sort; Qualitative research; Ultra-processed food.
Copyright © 2019. Published by Elsevier Ltd.