Development of DietSys: A comprehensive food and nutrient database for dietary surveys

Abstract

Background: Databases composed by nutrient composition tables, food groups, food processing categories, and servings and portion sizes adjusted for local population food habits or specific study aims are paramount to diet data assessment quality. Objective: We assembled a database, comprising the food composition tables of Brazil, Argentina and the USA, in which foods are classified into food groups, food processing categories, and assigned portion sizes to food items consumed locally, expanding the usability of the DietSys, a diet data processing system for the Brazilian population. Methods: Three food composition tables were incorporated, converted to an Excel® spreadsheet, coded, classified into food groups and classified into four food processing categories based on the NOVA classification system. Serving sizes from food photo albums and from tables were coded and assigned to food items. Results: The DietSys database comprises 9851 foods, from which 5901 were assigned of one to 27 servings or portion size options. Forty one percent of all items were classified as ultra-processed food products (UPP). Conclusion: The DietSys food and nutrient database will allow researchers to calculate food intake, food groups, food processing categories, and nutrient content of local foods representative of the Brazilian, Argentine, and North-American populations’ diet, particularly making it feasible the calculation of food intake from servings for the Brazilian population. The current iteration of DietSys is fully operational in Brazil, but to equally applicable to the USA and Argentinean populations, it requires including serving/portion sizes options representative of these countries.