Flavonoid intake and risk of Parkinson's disease

Background Flavonoids have been proposed to reduce the risk of Parkinson's disease (PD). However, results from epidemiological studies have been inconclusive. Objective To prospectively examine the association between the intake of flavonoids and their subclasses and the risk of PD and how pesticides may confound or modify that association. Methods The study population comprised 80 701 women (1984-2016) and 48 782 men (1986-2016) from two large US cohorts. Flavonoid intake was ascertained at baseline and every 4 years thereafter using a semiquantitative Food Frequency Questionnaire. We conducted multivariable-adjusted Cox regression models to estimate HRs and 95% CIs of PD according to quintiles of baseline and cumulative average intakes of flavonoids and subclasses. We repeated the analyses, adjusting for intakes of high-pesticide-residue fruits and vegetables (FVs) and stratifying by servings/day of high-pesticide-residue FV intake. Results We identified 676 incident PD cases in women and 714 in men after 30-32 years of follow-up. Higher total flavonoid intake at baseline was not associated with a lower PD risk, neither in men (HR comparing highest to lowest quintile: 0.89, 95% CI: 0.69 to 1.14) nor in women (HR comparing highest to lowest quintile: 1.27, 95% CI: 0.98 to 1.64). Similar results were observed for cumulative average intakes and flavonoid subclasses. Results remained similar after adjustment for and stratification by high-pesticide-residue FV and when analyses were restricted to younger PD cases. Conclusion These results do not support a protective effect of flavonoid intake on PD risk. Pesticide residues do not confound or modify the association.