Phenotype Refinement Influences GWAS-Implicated Variant Effect Sizes for Insomnia

Area of Science:

• Genetics
• Sleep Medicine
• Bioinformatics

Background:

• Genome-wide association studies (GWAS) are crucial for understanding complex traits but are limited by phenotype measurement, especially for sleep disorders.
• Subjective reports often conflate distinct sleep conditions like insomnia and restless legs syndrome (RLS), complicating genetic analyses.
• Improving phenotype classification is essential for advancing our understanding of the genetic architecture of sleep traits.

Purpose of the Study:

• To refine phenotypes for sleep traits using detailed questionnaire data from the UK Biobank.
• To differentiate between insomnia and RLS based on subjective reports in a large cohort.
• To re-evaluate significant genetic loci previously associated with insomnia.

Main Methods:

• Utilized expanded questionnaire data from approximately 180,000 UK Biobank participants.
• Employed statistical models to analyze phenotype data and genetic associations.
• Examined the influence of RLS on previously identified GWAS signals for insomnia at the MEIS1 and BTBD9 loci.

Main Results:

• Demonstrated that prior GWAS for insomnia likely mischaracterized participants due to a simplified, single-question approach.
• Found that two major GWAS signals for insomnia are predominantly or entirely driven by their association with RLS.
• Highlighted significant overlap and potential misattribution of genetic effects between insomnia and RLS.

Conclusions:

• Accurate phenotype classification is critical for the success of GWAS in sleep research.
• UK Biobank data offers valuable resources for detailed sleep phenotype refinement.
• Genetic associations previously attributed to insomnia may need re-evaluation due to confounding with RLS.