Reproducibility of a web-based FFQ for 13- to 15-year-old Danish adolescents

Abstract

FFQ are widely used in large-scale studies to assess dietary intake. To aid interpretation of diet–disease associations assessment of validity must be performed. Reproducibility is one aspect of validity focusing on the stability of repeated assessment with the same method which may also reveal problems in instrument design or participant instructions. The aim of the present study was to evaluate the reproducibility of a web-based FFQ targeting Danish adolescents within the Danish National Birth Cohort (DNBC). Data for the present study were obtained from a prospective design nested within the DNBC. Adolescents aged 13 to 15 years old (n 48, 60 % girls) completed the FFQ twice 4 weeks apart. The proportion of adolescents consistently classified into the same tertile according to amount of food intake ranged from 45 % (fish) to 77 % (vegetables), whereas classification into opposite tertiles ranged from 0 % (fruit, oils and dressing) to 15 % (beverages). Overall, no significant differences were observed in intake of food groups or nutrients between the two completions of the FFQ. Mean crude Spearman correlation for all food groups was 0·56 and mean intra-class correlation for all food groups was 0·61. In conclusion, the reproducibility of the FFQ for Danish adolescents was acceptable. The study revealed that adolescents aged 13–15 years seemed capable of recalling consistently overall dietary habits and had some difficulties estimating the frequency of consumption of regularly consumed food items.

Reproducibility of a Short-Form, Multicomponent Dietary Questionnaire to Assess Food Frequency Consumption, Nutrition Knowledge, and Lifestyle (SF-FFQ4PolishChildren) in Polish Children and Adolescents

“…These findings may indicate better test-retest reproducibility of the questionnaire in terms of dietary behaviors in children and adolescents with stable but restrictive eating habits (consuming key foods, healthy or unhealthy, with low frequency and/or consuming only a few selected foods with higher frequency). Better test-retest reproducibility of FFQs for foods consumed rarely or never has been reported in other studies [15,16,31]. The differences between both administrations of the questionnaire were more scattered in adolescents than in children, and the most between the adolescent's test and parent's test, which confirms the results of other statistical analyses.…”

“…The test-retest reproducibility for food items found in the present study in children (agreement: 59.0-98.4%, kappa: 0.46-0.81) and adolescents (42.9-80.1%, kappa: 0.30-0.44) was similar or higher than previously reported [31][32][33][34][35][36][37][38]. Parent-administered semi-quantitative FFQs to assess children's diet demonstrated cross-classification agreement, which ranged from 33-69% in Danish children [32] and 62.1-99.4% in Spanish children [37], but it ranged from 44-82% in New Zealand children using a short non-quantitative FFQ [35].…”

“…Lower reproducibility of food consumption frequency was found in children from six European countries (kappa: 0.23-0.68) [39]. For adolescents, the cross-classification agreement found in other reproducibility studies was similar or lower than our findings and ranged from: 29-58% in Italian adolescents [33], 36-55% in Norwegian adolescents [34], 37-87% in Belgian adolescents [36], 45-77% in Danish adolescents [31] and 46-88% in New Zealand adolescents [38], while the kappa values for foods obtained in adolescents were higher compared to our findings: 0.21-0.66 [33], 0.23-0.71 [31], 0.43-0.70 [36].…”

“…The present study showed lower reproducibility for more frequently consumed foods (juices, fruit, vegetables) and higher reproducibility for rarely consumed foods (energy drinks, fast food). Similarly, lower test-retest agreement was demonstrated for frequently consumed foods (e.g., white bread, wholemeal bread, fruit juices) and higher agreement for rarely consumed foods (e.g., porridge, fish, energy drinks) in Danish adolescents [31] and Polish adolescents and adults [15]. Estimating the consumption frequency of foods eaten less frequently or never could be easier than those eaten more often and included in various dishes and meals [11,31].…”

A Diet Score Assessing Norwegian Adolescents’ Adherence to Dietary Recommendations—Development and Test-Retest Reproducibility of the Score

“…Previous studies have generally reported lower percentage agreements in adolescents [38,39,40]. Øverby et al [38] included 58 Norwegian adolescents (14–15 years) who completed a 131-item FFQ four weeks apart.…”