Munich Cluster for Systems Neurology
print


Breadcrumb Navigation


Content

Reassessment of candidate gene studies for idiopathic restless legs syndrome in a large GWAS dataset of European ancestry.

Sleep. 2022 Apr 29:zsac098. doi: 10.1093/sleep/zsac098. Epub ahead of print. PMID: 35486972.

Authors/Editors: Schormair B, Zhao C, Salminen AV, Oexle K, Winkelmann J; International EU-RLS-GENE Consortium.
Publication Date: 2022

06_schormair

Abstract

Study objectives: Several candidate gene studies have been published for idiopathic restless legs syndrome (RLS) in populations of European ancestry, but the reported associations have not been confirmed in independent samples. Our aim was to reassess these findings in a large case-control dataset in order to evaluate their validity.

Methods: We screened PubMed for RLS candidate gene studies. We used the genome-wide association study (GWAS) dataset of the International-EU-RLS-GENE-Consortium as our replication sample, which provided genome-wide single variant association data based on at most 17,220 individuals of European ancestry. We performed additional gene-based tests using the software MAGMA and assessed the power of our study using the genpwr R package.

Results: We identified 14 studies conducted in European samples which assessed 45 variants in 27 genes of which five variants had been reported as significantly associated. None of these individual variants were replicated in our GWAS-based reassessment (nominal p > 0.05) and gene-based tests for the respective five genes ADH1B, GABRR3, HMOX1, MAOA, and VDR, were also non-significant (nominal p > 0.05). Our replication dataset was well-powered to detect the reported effects, even when adjusting for effect size overestimation due to winner's curse. Power estimates were close to 100% for all variants.

Conclusion: In summary, none of the significant single variant associations from candidate gene studies were confirmed in our GWAS dataset. Therefore, these associations were likely false-positive. Our observations emphasize the need for large sample sizes and stringent significance thresholds in future association studies for RLS.

Related Links