Validity of a semi-quantitative food frequency questionnaire based on food groups and cooking with a protein supplement in Japanese collegiate athletes

Abstract

A semi-quantitative food frequency questionnaire (FFQ) comprising a small number of food items may allow for more accurate assessment of habitual dietary intake in Japanese collegiate athletes. This study aimed to evaluate the validity and reproducibility of a commercially available FFQ (FFQg) with a newly added protein supplement item for assessing habitual dietary intake compared to the 24-hour dietary recall (24hDR) method using multiple statistical tests. A total of 108 participants (46 males) completed the FFQg on two separate days (FFQg1 and FFQg2) and 24hDRs on three non-consecutive days (two weekdays and one weekend day). Percentage differences, cross-classification analysis by quintile and Spearman’s rank correlation coefficients (CCs) were calculated to evaluate the validity and reproducibility of the FFQg. Results showed median percentage differences and de-attenuated CCs between the FFQg2 and 24hDRs were -3.4 (range: -46.8 - 25.1) and 0.33 (0.03 - 0.55), respectively, for 34 nutrients, and -16.1 (-51.4 - 45.0) and 0.34 (0.03 -0.64), respectively, for 18 food groups. Cross-classification analysis showed that median proportions of the “completely opposite category” between FFQg2 and 24hDRs for energy-adjusted intakes of nutrients and food groups were both within 5%. Median CCs for energy-adjusted intake between the FFQ1 and FFQ2 were 0.61 (0.40 - 0.82) for nutrients and 0.52 (0.40 - 0.87) for food groups. In conclusion, although estimating habitual dietary intake requires careful consideration, the FFQg offers reasonably valid measurements of ranking individuals to assess intakes of macronutrients and food groups in collegiate athletes, with a relatively good reproducibility.