Design
Prospective epidemiologic studies published before January 2018 were searched through 2 databases. Log-transformed multivariable adjusted hazard and odds ratios were combined in a random-effects model. Meta-analyses comparing extreme quantiles of polyphenol exposure were further explored with the use of linear and nonlinear dose-response meta-analyses.
Results
Eighteen studies investigated the association between polyphenols (51 different compounds in total) and type 2 diabetes. A comparison of extreme quantiles revealed inverse associations for intakes of polyphenols (HR: 0.56; 95% CI: 0.34, 0.93), flavonoids (HR: 0.88; 95% CI: 0.81, 0.96), flavonols (HR: 0.92; 95% CI: 0.85, 0.98), flavan-3-ols (HR: 0.89; 95% CI: 0.81, 0.99), catechins (HR: 0.86; 95% CI: 0.75, 0.97), anthocyanidins (HR: 0.86; 95% CI: 0.81, 0.91), isoflavones (HR: 0.92; 0.86, 0.97), daidzein (HR: 0.89; 95% CI: 0.83, 0.95), genistein (HR: 0.92; 95% CI: 0.86, 0.99), and stilbenes (HR: 0.44; 95% CI: 0.26, 0.72), and biomarkers of daidzein (HR: 0.81; 95% CI: 0.66, 0.99) and genistein (HR: 0.79; 95% CI: 0.62, 0.99). In the dose-response meta-analysis, nonlinear associations were observed for intakes of polyphenols, flavonoids, flavanones, anthocyanidins, anthocyanins, and biomarkers of genistein. A linear dose-response association was observed for phenolic acids.
Conclusions
This study adds to the evidence showing that diets rich in polyphenols, and particularly flavonoids, play a role in the prevention of type 2 diabetes. For most associations evidence for nonlinearity was found, suggesting a recommendable amount of intake associated with the lowest risk of type 2 diabetes. Therefore, future studies are warranted in which nonlinear associations are further explored.
