Food frequency questionnaire for Japanese athletes: validation with junior high school athletes

To investigate the validity of a food frequency questionnaire (FFQ) developed for Japanese athletes (FFQJA) among junior high school students.

Sixty-four junior high school soccer players (13-15 years old) completed the FFQ and provided dietary records (DRs) for 3 nonconsecutive days (2 weekdays and 1 weekend day). Individual ranking estimates, crude Spearman's correlation coefficients, and energy-adjusted correlation coefficients were used to validate the FFQ. To evaluate the mean estimates, the differences and percentage differences between the results obtained from the DRs and FFQ were calculated. The consistency of intakes between the two methods and the validity of classification were evaluated.

The median crude correlation coefficients between the questionnaire responses and DRs were 0.483 (range: 0.350-0.708) and 0.386 (range: 0.191-0.684) for nutrients and food groups, respectively. The median energy-adjusted correlation coefficients were 0.477 (range: 0.340-0.722) and 0.384 (range: -0.112 to 0.744), respectively. The median percentage differences were -7.2 and -17.6, respectively, and the median intraclass correlation coefficients were 0.469 and 0.214, respectively. Cross-classification analysis showed that the median kappa values were 0.350 and 0.275, respectively. The misclassification rate was 3% for nutrients and 6% for food groups.

The FFQ developed and validated for Japanese athletes was found to be valid for Japanese junior high school soccer players. Therefore, the FFQJA could be a suitable tool for ranking and categorizing individuals among Japanese junior high school athletes.