Validation of a food frequency questionnaire for individuals undergoing colorectal cancer screening

The objective of this study was to validate the Food Frequency Questionnaire (FFQ) for the adult and elderly population in colorectal cancer screening in the Legal Amazon region. The FFQ and two 24-hour recalls were applied, and nutrient intake estimates were compared. For this, the paired t test, Pearson correlation and Bland-Altman regression were used with a significance level of 5%. Significant correlations were accepted, with R values between 0.4 and 0.7 for macronutrients and 0.3 to 0.7 for micronutrients. For energy, carbohydrates, total and insoluble fiber, lipids, omega-3, monounsaturated, polyunsaturated, saturated fat, calcium, iron, zinc, retinol, cholecalciferol and ascorbic acid, a positive correlation was observed between the methods, being considered acceptable in the analysis of validation. Only for soluble fiber, selenium, and proteins the R values obtained were lower than 0.30. Bland-Altman analyzes showed that the methods correlate best when daily energy levels are ingested within the patient's needs. This FFQ is considered a good tool to be used in epidemiological studies to assess the food intake of these patients. However, caution is advised when using it with individuals with energy intakes above or below recommended daily levels.