Development and validation of a food frequency questionnaire for dietary intake assessment among multi-ethic primary school-aged children

Abstract

INTRODUCTIONThis study aimed to develop and validate a food frequency questionnaire (FFQ) to assess habitual diets of multi-ethnic Malaysian children aged 7-12 years.METHODS A total of 236 primary school children participated in the development of the FFQ and 209 subjects participated in the validation study, with a subsample of 30 subjects participating in the reproducibility study. The FFQ, consisting of 94 food items from 12 food groups, was compared with a three-day dietary record (3DR) as the reference method. The reproducibility of the FFQ was assessed through repeat administration (FFQ2), seven days after the fi rst administration (FFQ1). RESULTSThe results of the validation study demonstrated good acceptance of the FFQ. Mean intake of macronutrients in FFQ1 and 3DR correlated well, although the FFQ intake data tended to be higher. Cross-classifi cation of nutrient intake between the two methods showed that < 7% of subjects were grossly misclassifi ed. Moderate correlations noted between the two methods ranged from r = 0.310 (p < 0.001) for fat to r = 0.497 (p < 0.001) for energy. The reproducibility of the FFQ, as assessed by Cronbach's alpha, ranged from 0.61 (protein) to 0.70 (energy, carbohydrates and fat). Spearman's correlations between FFQ1 and FFQ2 ranged from rho = 0.333 (p = 0.072) for protein to rho = 0.479 (p < 0.01) for fat.CONCLUSION These fi ndings indicate that the FFQ is valid and reliable for measuring the average intake of energy and macronutrients in a population of multi-ethnic children aged 7-12 years in Malaysia.

Relative validity of a semi-quantitative food frequency questionnaire for Singaporean toddlers aged 15–36 months

“…In these analyses, it was found that total fat, SFA, MUFA, vitamin A and E had the weakest correlations and/or agreements between methods. Again, this was consistent with the literature, particularly for fats [26,27,48,49]. There may be several reasons for this.…”

“…Results indicated that overall, the sqFFQ overestimated intakes of all nutrients when compared to the WFR. This finding was consistent with literature and most likely attributable to the format of the sqFFQ [26,27,48,49]. With the traditional format of a sqFFQ, not only did parents have to think retrospectively about their child's habitual intake, but they also had to consider frequency of intake, based on the portion size presented next to each food.…”

“…Reproducibility was assessed by asking participants to complete a second sqFFQ. As there were no guidelines indicating whether the full sample was needed for reproducibility, or, a proportion of the sample was sufficient, 20 % of the sample were asked to complete a second sqFFQ one to two weeks later [27,34,36].…”

“…The age ranges included in these studies were 1.5-4 years. Even fewer FFQs are available for children in Asia [27,28], of which none involve children less than three years of age. Therefore, presently, there is no simple tool for use in large epidemiological studies to understand the food and nutrient intakes of Asian toddlers.…”

Validation of ‘POIBA-How do we eat?’ questionnaire in 9-10 years old schoolchildren

“…We analysed the internal consistency and reliability of POIBA-HDWE FFQ for the food frequency questions using Cronbach’s alpha, which was expected to be 0.6–0.7 [21,22]. Additionally, the validity of the POIBA-HDWE FFQ was evaluated by analysing the sensitivity, specificity, positive predictive values (PPV) and negative predictive values (NPV) with the variable ‘Food consumers’.…”