Reproducibility and validity of a simple checklist-type questionnaire for food intake and dietary behavior

Abstract

BACKGROUND: A simple, reliable, and valid food questionnaire is needed in clinical dietary assessments, community health education, and multi-purpose epidemiologic studies to obtain a crude measure of dietary intake. METHODS: To assess the validity and reproducibility of a simple 4-point scale food intake and behavior checklist, it was compared to two 3-day weighed dietary records. The FBC was administered to 47 students of a dietician course and their parents (n=94) over a 9-month interval to assess the reproducibility. The mean intakes of selected food groups assessed by the two dietary records completed between food intake and behavior checklists were compared to the responses to the food intake and behavior checklist to assess its validity. RESULTS: The kappa statistics for reproducibility ranged from 0.25 for confectionaries to 0.63 for a preference for fatty foods (median, 0.39). There was a reasonable level of correlation between the dietary record and the food intake and behavior checklist in the intake of eggs, milk, and fruits (r=0.53, 0.56, and 0.50, respectively). There was a weaker but still significant correlation in the intake of vegetables, and alcohol (r=0.31 and 0.45, respectively). No significant correlation was observed in the intake of meat, fish, confectionaries, and soft drinks. However, those who reported consuming mainly fish rather than meat were found to eat significantly less meat and animal fat. Similarly, those who did not prefer fatty foods consumed significantly less meat, animal fat, and polyunsaturated fatty acids. CONCLUSIONS: This simple food checklist was useful in collecting data on egg, milk, and fruit consumption. Assessing intake frequency of vegetables, meat or fish with the FBC may be useful in screening high- or low-intake individuals.