Validation and Calibration of a Semiquantitative Food Frequency Questionnaire Designed for Adolescents

Background Epidemiologic studies must have a reliable method for evaluating food intake; therefore, valid, precise, and practical instruments are essential. Objective To assess the relative validity and estimate the calibration factors of a semiquantitative food frequency questionnaire (FFQ) for adolescents. Design Validation and calibration study. Subjects/settings This study enrolled 169 adolescents from Rio de Janeiro, Brazil. A set of three food records was used as the reference method. Statistical analyses performed The relative validity was analyzed according to weighted kappa values for the quartile categorization of energy and nutrient intakes, Pearson correlation coefficients, and the Bland-Altman method. To estimate the calibration factors for the FFQ, linear regression models, including the food record means as dependent variables and the FFQ estimations as independent variables were developed for boys and girls. Results The weighted kappa values ranged from 0.28 to 0.44 for the raw data and from 0.16 to 0.39 for the deattenuated and energy-adjusted data. The Pearson correlation coefficients ranged from 0.33 to 0.46, and the mean agreement varied from 62% to 143%. The calibration factors estimated for boys ranged from 0.15 to 0.48, and the factors estimated for girls ranged from 0.14 to 0.47. The mean energy and nutrient intakes estimated by the calibrated FFQ were similar to the means estimated by the food records; however, the standard deviations were smaller for the calibrated FFQ estimations. Conclusions The tested FFQ is a suitable tool for ranking energy and nutrients intake in the studied group. Calibration factors are needed to estimate energy and nutrient intake means and should be used to correct raw data as well as association measurements based on FFQ data. J Am Diet Assoc. 2010;110:1170-1177.