Validation of the Alzheimer’s disease-Commonwealth Scientific and Industrial Research Organisation food frequency questionnaire using weighed food records and biomarkers: the Method of Triads model

Abstract The objective was to apply the triangular approach, the Method of Triads, to determine the validity of the Alzheimer’s disease-Commonwealth Scientific and Industrial Research Organisation food frequency questionnaires (FFQ) in 98 cognitively unimpaired Australian older adults using blood biomarkers and 4-day weighed food records. Spearman’s correlation coefficients were calculated between each dietary method and the validity coefficient was calculated using the Method of Triads. 95% confidence intervals for the validity coefficients were estimated using bootstrap sampling. Spearman’s correlation coefficients between FFQ and weighed food record intake for docosahexaenoic acid (DHA) and vitamin E (r = 1.00 and r = 0.95, respectively), and the FFQ intake and β carotene biomarker level were positively associated (r = 0.35; all p < .001). The validity coefficient of the FFQ was highest for DHA (0.94), with a moderate coefficient observed for β carotene (0.47). The Method of Triads could not be applied to vitamin E and eicosapentaenoic acid as one of the individual correlation coefficients was negative. Validity coefficients were comparable to those of the few other published studies utilizing this same approach. The study provides evidence towards validity of the Alzheimer’s disease-Commonwealth Scientific and Industrial Research Organisation FFQ in the measurement of DHA and β carotene, however, was unable to determine validity of vitamin E and eicosapentaenoic acid intake using this method.