Reproducibility and validity of a diet quality index for children assessed using a FFQ

Abstract

The diet quality index (DQI) for preschool children is a new index developed to reflect compliance with four main food-based dietary guidelines for preschool children in Flanders. The present study investigates: (1) the validity of this index by comparing DQI scores for preschool children with nutrient intakes, both of which were derived from 3 d estimated diet records; (2) the reproducibility of the DQI for preschoolers based on a parentally reported forty-seven-item FFQ DQI, which was repeated after 5 weeks; (3) the relative validity of the FFQ DQI with 3 d record DQI scores as reference. The study sample included 510 and 58 preschoolers (2.5-6.5 years) for validity and reproducibility analyses, respectively. Increasing 3 d record DQI scores were associated with decreasing consumption of added sugars, and increasing intakes of fibre, water, Ca and many micronutrients. Mean FFQ DQI test-retest scores were not significantly different: 72 (sd 11) v. 71 (sd 10) (P = 0.218) out of a maximum of 100. Mean 3 d record DQI score (66 (sd 10)) was significantly lower than mean FFQ DQI (71 (sd 10); P < 0.001). The reproducibility correlation was 0.88. Pearsons correlation (adjusted for within-person variability) between FFQ and 3 d record DQI scores was 0.82. Cross-classification analysis of the FFQ and 3 d record DQI classified 60 % of the subjects in the same category and 3 % in extreme tertiles. Cross-classification of repeated administrations classified 62 % of the subjects in the same category and 3 % in extreme categories. The FFQ-based DQI approach compared well with the 3 d record approach, and it can be used to determine diet quality among preschoolers.