Development and evaluation of a food frequency questionnaire for use among young children

Abstract

Background/objectives: This study described the development of a parent food frequency questionnaire (FFQ) for measuring diets of young children over the past month and the validation of this FFQ against three non-consecutive 24 hour recalls.

Subjects/methods: Food and nutrient intakes from a 68-item FFQ were compared with three non-consecutive 24 hour recalls in a follow-up cohort of children aged 1.5, 3.5 and 5.0 years old. Data from both methods were available for 231, 172 and 187 participants at ages 1.5, 3.5 and 5.0 years, respectively.

Results: Out of 11 nutrients, four (protein, fat, fibre, iron), two (Vitamin C, folate) and three (protein, vitamin C and folate) nutrients showed good-acceptable outcome for 2 out of 3 group-level validation tests at ages 1.5, 3.5 and 5.0 years, respectively. Of 26 food groups, good-acceptable outcome for 2 out of 3 group-level validation tests was revealed for two, four and six food groups at ages 1.5, 3.5 and 5.0 years, respectively. For individual-level validation tests, all nutrients showed good-acceptable outcome for 2 out of 3 individual level tests across three time points, except for folate at age 1.5 years and energy intake at age 3.5 years. Most food groups (22 out of 26) at age 1.5 years and all food groups at both ages 3.5 and 5.0 years showed good-acceptable outcome for 2 out of 3 individual-level validation tests.

Conclusions: At all three time points, the FFQ demonstrated good-acceptable validity for some nutrients and food groups at group-level, and good-acceptable validity for most nutrients and food groups at individual-level. This quantitative FFQ is a valid and robust tool for assessing total diet of young children and ranking individuals according to nutrient and food intakes.

Fig 1. Flowchart of recruitment and exclusion of participants in the Melbourne InFANT program for FFQ validation and analyses.