Reproducibility and comparison of a digital food frequency questionnaire (DIGIKOST-FFQ) assessing adherence to national diet and lifestyle recommendations.

Background: We have developed a digital semi-quantitative food frequency and lifestyle questionnaire, the DIGIKOST-FFQ, based on the validated paper-based NORDIET-FFQ.

Objective: The study aims to investigate the reproducibility of the DIGIKOST-FFQ and to compare the DIGIKOST-FFQ against the NORDIET-FFQ for the adjusted questions for intakes of fruits, vegetables, whole grains, fish, meat, and dairy products.

Design: Participants were recruited from May to September 2021 through a random sample from the National Population Register and advertisements on Facebook in Norway. In the reproducibility study, the DIGIKOST-FFQ was completed twice by the participants, 1-2 months apart. In the comparison study, the DIGIKOST-FFQ was completed 1-2 months prior to the NORDIET-FFQ.

Results: In the reproducibility study, 317 individuals were included. For 12 out of 16 food groups there were no significant differences in intake estimations between the first and second DIGIKOST-FFQ administrations. A small but significant median difference was observed for fruits (6 g/day) and vegetables (24 g/day). Correlations were satisfactory for all items (r = 0.60-1.00), and in the cross-classification 85% of the participants were classified into the same or adjacent quartile for all items. The comparison study included 81 individuals. Compared to the NORDIET-FFQ a significant median difference was observed for fruits 29 g/day, vegetables 36 g/day, whole grains -10 g/day, and red meat -11 g/day, but not for fish, processed meat, or dairy products.

Conclusion: The DIGIKOST-FFQ was able to reproduce diet and lifestyle at the group level. An intended difference for the food groups where questions had been adjusted, was observed between DIGIKOST-FFQ and NORDIET-FFQ in the comparison study.