Development and validation of a food-frequency questionnaire to assess short-term antioxidant intake in athletes.

A quantitative food-frequency questionnaire (FFQ) was developed to determine antioxidant intake in athletes. The questionnaire will be valuable for researchers wishing to standardize antioxidant intake or simply document habitual intake during an intervention trial. One hundred thirteen athletes participated in the validity study, of whom 96 completed the questionnaire and blood test, 81 completed the 7-d food diary and questionnaire, and 63 completed the 7-d food diary and blood test. Validity was investigated by comparing total and food-group antioxidant intakes from the questionnaire with those from a subsequent 7-d food diary. Measures of construct validity were determined by comparing a biomarker of antioxidant capacity (ferric-reducing ability of plasma) in a blood sample with antioxidant intakes from the questionnaire and diary. The correlation between the diary and questionnaire energy-adjusted estimates of total antioxidant intake was modest (.38; 90% confidence limits, ± .14); the correlation was highest for antioxidants from cereals (.55; ± .11), which contributed the greatest proportion (31%) of the total antioxidant intake. Correlations were also high for coffee and tea (.51; ± .15) and moderate for vegetables (.34; ± .16) and fruit (.31; ± .16). The correlation of the plasma biomarker with the questionnaire estimate was small (.28; ± .15), but the correlation with the diary estimate was inconsequential (-.03; ± .15). One-week test-retest reliability of the questionnaire's estimates of antioxidant intake in 20 participants was high (.83; ± .16). In conclusion, the FFQ is less labor intensive for participants and researchers than a 7-d diary and appears to be at least as trustworthy for estimating antioxidant intake.