4.4. Dietary Assessment Methods
In this study, adolescents rated their own intake of fruits and vegetables differently using the short food behavior questionnaire compared to how they recorded their intake using a three day food record. Despite the reducing levels of adherence and modest changes in diet, as shown by the food records, the estimates of fruit and vegetable intake from the eating behavior questionnaire remained significantly increased following intervention for the entire 12 month maintenance period. For example, the food records showed no change in vegetable intake following intervention but the short questionnaire showed an increase of vegetable intake during the maintenance period of up to 0.9 serves at 12 months post-intervention. Thus, it seems that the food behavior questionnaire may have overestimated the effect of the intervention on intakes as compared to the food records, particularly for vegetables. This discrepancy may be due to a desire to report socially acceptable intakes in line with the CAFAP key messages [52], particularly given that all changes occurred after the intervention had been delivered. The wording of the questions directly reflected the nutrition intervention messages, so adolescents may have felt obliged to show they had adhered. Alternatively the participants may have truly believed that they were eating more healthfully. The differences in self-reported behaviors from questionnaire and self-reported intake from food records highlight the inherent difficulties in obtaining consistent and accurate nutrition data in this population.