BackgroundThe purpose of the study was to examine the association of long and short sleep duration with risk of Parkinson’s disease (PD) across RORA rs2028122 genotypes.MethodsIn the present prospective study with a large sized UK Biobank cohort, we performed multivariate logistic regression analyses, generalized additive model, interaction terms, stratification analysis, and mediation analysis to evaluate the association of long and short sleep duration with risk of PD across RORA rs2028122 genotypes.ResultsThe GG genotype [1.16 (1.01, 1.33)], a short sleep duration [1.23 (1.10, 1.37)], and a long sleep duration [1.19 (1.03, 1.37)] were identified as the independent risk factors for PD. Sleep duration exhibited a curvilinear U-shaped correlation with the risk of PD; first, the risk of PD gradually decreased as the length of sleep increase, but then, the risk began to increase as the length of sleep increase. Among habitual long sleepers, AG carriers had a higher risk of PD compared with AA carriers [1.67 (1.09, 2.55)]. Among AG carriers, both habitual short [1.28 (1.09, 1.50)] and long [1.38 (1.13, 1.69)] sleepers increased the risk of PD compared with habitual normal sleepers. Among GG carriers, habitual short sleepers have a higher risk of PD [1.26 (1.06, 1.50)] compared with habitual normal sleepers. A mediation model suggested that the rs2028122 genotype partially mediated the causal pathway of sleep duration leading to the development of PD on a positive effect.ConclusionOur study demonstrated that the association between sleep duration and PD risk varied across different RORA rs2028122 genotypes. Our findings could help individuals to identify their potential risk profile and take timely actions to prevent the PD.