OBJECTIVE: This study sought to develop an instrument to obtain and assess data on the qualitative and quantitative food intake of population groups. METHODS: Initially, data on food intake of a representative sample of inhabitants of Viçosa (n=119) were collected. The 24-hour recall and direct weight were used to determine the most consumed foods and their portions. Interview was done at lunch or dinner where all foods and portions mentioned in the previous 24 hours were weighed; a food picture album was used to aid in determining the weight of the portion. RESULTS: From the 60 most frequently consumed foods and the identified portions, a questionnaire based on the semi-quantitative Food Intake Frequency method, with 65 items and a food picture album were developed. A pilot study was done to assess this methodology in a subsample (n=34). To compare the two methods, the energy value and macronutrients were assessed, using the Pearson coefficient and the Student's "t" test and the values obtained were r=0.53 for energy, r=0.43 for proteins, r=0.58 for lipids and r=0.21 for carbohydrates. A significant difference was not detected by the Student's "t" test for proteins and lipids between the two methods (alpha=0.05). CONCLUSION: The questionnaire was proven to be a promising method for population surveys, yet more studies are necessary to determine its limitations and possibly expand the questionnaire in order to cover the entire population.
nutrition assessment; food consumption; eating; nutrition survey
