The association of dietary total flavonoids and their subclasses with the risk of type 2 diabetes: a prospective cohort study

Background Data from mechanistic studies suggest flavonoids may benefit glucose metabolism, but their associations with type 2 diabetes (T2D) remain unclear. This study examined the prospective associations of dietary intake of total, classes, and individual flavonoids, as well as their source foods, with T2D in the CArdioVascular disease Association Study (CAVAS). Methods A total of 16,666 Korean men and women were enrolled at baseline, and 953 were newly diagnosed with T2D over a median follow-up of 5.96 years. Intake of flavonoids was cumulatively averaged using all food frequency questionnaires before the censoring events. A Poisson regression model was used to estimate incidence rate ratios (IRRs) and 95% confidence intervals (CIs). Results Women with higher total flavonoid, flavonol, isoflavone, and proanthocyanidin intake had a lower risk of T2D (fourth vs. first quartile, IRR 0.62; 95% CI 0.44-0.89; P for linearity and non-linearity < 0.05 for total flavonoids), while in men, flavanones, anthocyanins, and proanthocyanidins, but not total flavonoids, were inversely associated with T2D risk (all P interaction for sex > 0.05). The key source foods contributing to flavonoid intake were also different between men and women, except for apples: tangerines and strawberries in men and green leafy vegetables and soy products in women. Conclusions A higher intake of total flavonoids, particularly from vegetables, soybeans, and apples, may be associated with lower risk of T2D in women. However, flavonoids from fruits, rather than total flavonoids, may be inversely associated in men. The association between flavonoid intake and the risk of T2D may be contingent upon the dietary sources of flavonoids consumed.