OBJECTIVE: To verify the validity of the Food Frequency Questionnaire for Adolescents in order to evaluate food group consumption among adolescents from Piracicaba, São Paulo. METHODS: The study was carried out with 94 adolescents, ranging from 11 to 15 years of age, from the public education system. Food consumption was assessed by the Food Frequency Questionnaire for Adolescents (QFAA) and the average of two 24-hour recalls (R24h) was used as a reference method. The food items were classified into 18 groups. Descriptive analyses, paired t-test and Wilcoxon test, Pearson and Spearman coefficients were performed. Analysis of quartiles and Kappa test were used. Correlation coefficients were adjusted by the within-subject variance of the 24-hour recalls, from one-way ANOVA. RESULTS: No significant differences were found between the tools for rice, pasta, meat, soft drinks, and artificial juices. The correlation coefficients adjusted by within-subject variability ranged from -0.26 to 0.78. Agreement of individuals' classification in quartiles of food consumption for both methods ranged from 22% (pasta) to 50% (beans). For discordant quartiles, the groups that presented more than 10% of misclassified subjects were pasta (19%), meat (13%) and fats (11%). The values of the Kappa test ranged from -0.15 (pastas) to 0.56 (beans). The food frequency questionnaire over-estimated the consumption of almost all the food groups and under-estimated the groups of oils, beans, meats, and soft drinks CONCLUSION: The food frequency questionnaire revealed good validity for beans, vegetables, milk and dairy, cookies with filling, and rice.
Adolescents; Dietary assessment; Food groups; Food frequency questionnaire; Validity
