A food frequency questionnaire validated for estimating dietary flavonoid intake in an Australian population

Abstract

Flavonoids, a broad category of nonnutrient food components, are potential protective dietary factors in the etiology of some cancers. However, previous epidemiological studies showing associations between flavonoid intake and cancer risk have used unvalidated intake assessment methods. A 62-item food frequency questionnaire (FFQ) based on usual intake of a representative Australian adult population sample was validated against a 3-day diet diary method in 60 young adults. Spearman's rank correlations showed 17 of 25 individual flavonoids, 3 of 5 flavonoid subgroups, and total flavonoids having strong/moderate correlation coefficients (0.40-0.70), and 8 of 25 individual flavonoids and 2 of 5 flavonoid subgroups having weak/insignificant correlations (0.01-0.39) between the 2 methods. Bland-Altman plots showed most subjects within ±1.96 SD for intakes of flavonoid subgroups and total flavonoids. The FFQ classified 73-90% of participants for all flavonoids except isorhamnetin, cyanidin, delphinidin, peonidin, and pelargonidin; 73.3-85.0% for all flavonoid subgroups except Anthocyanidins; and 86.7% for total flavonoid intake in the same/adjacent quartile determined by the 3-day diary. Weighted kappa values ranged from 0.00 (Isorhamnetin, Pelargonidin) to 0.60 (Myricetin) and were statistically significant for 18 of 25 individual flavonoids, 3 of 5 subgroups, and total flavonoids. This FFQ provides a simple and inexpensive means to estimate total flavonoid and flavonoid subgroup intake.