OBJECTIVE: The aim of the present study was to investigate the validity and reproducibility of a food frequency questionnaire developed for the adult population. METHODS: Sixty-nine individuals from both genders who attended industrial restaurants from two government divisions in Brasília, Federal District, Brazil, were interviewed on three different occasions. In the first and third interviews, a 24h recall and a food frequency questionnaire were applied, and in the second interview only a 24h recall was applied. The study lasted 181 days. Macronutrients, vitamins A and C, calcium, iron, zinc, cholesterol, fibers and total energy were analyzed using the intraclass and Pearson's correlation coefficients, this last one in a crude and deattenuated manner. RESULTS: Regarding reproducibility, the food frequency questionnaire provided adequate results, presenting correlation coefficients above 0.7 for lipid, cholesterol, energy and protein. For validity, the best deattenuated coefficients were observed for vitamin C (r=0.66), iron (r=0.58), protein (r=0.55) and carbohydrate (r=0.55) and the worst for cholesterol (r=0.32) and vitamin A (r=0.37). CONCLUSION: The expressive correlation coefficients obtained for some nutrients indicates that this food frequency questionnaire constitutes a good research instrument for epidemiological studies in the adult population, as it can supply important dietary information allowing the implementation of actions in the health and nutrition areas.
food consumption; nutrition surveys; 24-hour recall; reproducibility of results
