Abstract
Parkinson's disease (PD) and restless legs syndrome (RLS) may be clinically and/or etiologically related, yet this association is under debate. Single-nucleotide polymorphisms (SNPs) in the TOX3 gene locus were implicated in both RLS and PD genome-wide association studies (GWASs), suggesting a potential pleiotropy. Two case-control cohorts including 644 PD patients, 457 RLS patients, and 945 controls were genotyped for one known RLS-related SNP (rs3104767) and one PD-related SNP (rs4784226) in the TOX3 locus. The associations between genotype and PD and RLS risk were tested using multivariate regression models. The allele frequencies of RLS-related SNP rs3104767 in RLS patients and controls were 0.35 and 0.43, respectively (OR 0.70, p = 0.0007). Regression model suggested that this association is derived by homozygous carriage of rs3104767 (adjusted p = 0.008). A nominal association was observed for homozygous carriers of the rs3104767 SNP in PD (OR 1.62, 95% CI 1.05-2.54, p = 0.034), i.e., with an opposite direction of effect on RLS and PD, but this was not significant after Bonferroni correction. However, data from published GWASs of RLS and PD, and from the PDgene database, further supported these inverse associations. Our results confirm the association between the TOX3 SNP rs3104767 and RLS and suggest that TOX3 variants are involved in both RLS and PD, but with different or even opposite effects. Studies in larger populations of different ethnicities are required to further refine the TOX3 locus is involved in RLS and PD.