Development, validity and reproducibility of a whole grain food frequency questionnaire in Malaysian children

Background To date, there is no validated whole grain assessment tool for children in any Southeast Asian countries. Hence, there is a need for a valid tool to assess whole grain intake among Malaysian children. This study aimed to develop, validate and test the reproducibility of a food frequency questionnaire (FFQ) in estimating whole grain intake among Malaysian children. Methods A total of 392 children participated in the FFQ development and 112 children aged 9-12 years participated in the validation phase; with a subsample of 50 children participating in the reproducibility phase. Three-day diet record (3DR) as the reference method in validation phase. Spearman correlations, mean difference, Bland-Altman plot and cross-classification analyses were used to assess validity. The reproducibility was tested through a repeat administration of the FFQ, with 1 month time interval. Reproducibility analyses involved intra-class correlation coefficient (ICC), Cronbach's alpha and cross-classification analyses. Results The FFQ consisted of 156 whole grain food items from six food groups. Mean intake of whole grain in FFQ1 and 3DR were correlated well (r = 0.732), demonstrated good acceptance of the FFQ. Bland Altman plots showed relatively good agreement for both the dietary methods. Cross-classification of whole grain intake between the two methods showed that < 9.9% of children were grossly misclassified. Outcomes from ICC (0.989) and Cronbach's alpha (0.995) demonstrated excellent reliability. All the children were classified in the same or adjacent quartile of whole grain intake. Conclusions Overall, the findings support the validity of the developed FFQ to appropriately estimate the whole grain intake in Malaysian children. This validated FFQ will be a valuable tool for future studies, to analyses the impact of whole grain consumption with disease relationship among Malaysian schoolchildren.