Relative impact of flavonoid composition, dose and structure on vascular function: A systematic review of randomised controlled trials of flavonoid-rich food products

Scope Previous systematic reviews suggest beneficial effects of flavonoids on biomarkers of cardiovascular disease (CVD) risk, but have overlooked the impact of dose response or food complexity. The aim of the present study was to examine the relative impact of composition, flavonoid structure and dose. Methods and results MEDLINE, EMBASE and Cochrane were searched for randomised controlled trials (RCTs) of flavonoids or flavonoid‐rich foods/extracts. Flavonoid composition was established using United States Department of Agriculture (USDA) and Phenol‐Explorer databases. Effects of six flavonoid subgroups on endothelial function (flow‐mediated dilation; FMD), and systolic and diastolic blood pressures were assessed by random effects meta‐analyses and regression analyses. Meta‐analyses of combined flavonoid subclasses showed significant improvements in FMD (chronic, 0.73% (0.17, 1.30) 14 RCTs; acute, 2.33% (1.58, 3.08) 18 RCTs) and blood pressures (systolic, –1.46 mmHg (–2.38, –0.53) 63 RCTs; diastolic, –1.25 mmHg (–1.82, –0.67) 63 RCTs). Similar benefits were observed for the flavan‐3‐ol, catechol flavonoids (catechins, quercetin, cyanidin etc.), procyanidins, epicatechin and catechin subgroups. Dose‐response relationships were non‐linear for FMD (R2 ≤ 0.30), with greater associations observed when applying polynomial regression analyses (R2 ≤ 0.72); there was no indication of a dose response for blood pressure. Conclusion The present analysis suggests that flavonoid bioactivity does not follow a classical linear dose‐response association and this may have important biological implications.