Relative Validity of Nutrient and Food Group Intakes Estimated From the Brief-type Self-administered Dietary History Questionnaire for Japanese Preschool Children Aged 3-6 Years (BDHQ3y).

Abstract

Background: To evaluate the relative validity of nutrient and food group intakes estimated using the brief-type self-administered Diet History Questionnaire for Japanese preschool children aged 3-6 years (BDHQ3y).

Methods: This study included 500 children (253 boys and 247 girls) aged 3-6 years recruited from 32 prefectures in Japan. Guardians completed the BDHQ3y to report their child's dietary intake over the previous month and then 8-day weighed dietary records (DRs) over a year (2 days per season). From the BDHQ3y and DRs, energy-adjusted intakes of 44 nutrients and 26 food groups were estimated based on the density method.

Results: Percentage differences in median intakes were within 10% for 26 nutrients (most energy-providing nutrients and several vitamins and minerals) and three food groups (cereals, fish/shellfish and milk/dairy products). The median Spearman's rank correlation coefficients were 0.31 (interquartile range [IQR], 0.23-0.40) for nutrients and 0.32 (IQR, 0.23-0.41) for food groups. Correlations exceeded 0.40 for 11 nutrients (eg, added sugar, dietary fiber and calcium) and nine food groups (eg, vegetables, fruits and milk/dairy products). Bland-Altman plots showed wide limits of agreement, with overestimation of intakes of protein, fats, carbohydrates, fruits, and milk/dairy products, as well as underestimation of intake of vegetables at higher intake levels. Similar findings were made when boys and girls were analyzed separately.

Conclusion: The BDHQ3y has the potential to monitor population-level intakes of a range of nutrients. However, its utility for ranking individuals and assessing individual-level intake remains limited, requiring cautious interpretation and application.