Polygenic risk scores and Mendelian randomization reveal circadian genetic contributions to idiopathic hypersomnia

Abstract Idiopathic hypersomnia (IH) is a rare and heterogeneous sleep disorder characterized by excessive daytime sleepiness. We aimed to stratify IH patients based on polygenic risk scores (PRSs) for sleep-related traits and explored the underlying genetic predispositions. Genome-wide single-nucleotide polymorphism data from 303 Japanese IH patients and 2,918 controls were analyzed. PRSs were calculated for chronotype (morningness/eveningness), daytime napping, sleep duration, and insomnia using publicly available base data. Patients in the extreme PRS tails were examined for clinical phenotypes. Causal inference was evaluated via Mendelian randomization (MR). IH patients were significantly enriched in the top PRS percentiles for eveningness (top 0.5%: odds ratio [OR] = 3.40, P = 3.9×10 −5 ; top 1%: OR = 2.48, P = 1.9×10 −4 ) and daytime napping (top 0.5%: OR = 3.80; P = 1.6×10 −5 ), with substantially elevated ORs. Patients with high morningness PRS exhibited increased slow-wave sleep near wake time, frequently observed in patients with IH. MR analysis supported a causal relationship between eveningness and IH (inverse-variance weighted, UK Biobank: P = 1.3×10 −3 ; replicated in 23andMe P = 0.037). This causal association was consistently observed even among a subgroup of IH patients who showed ≥660 minutes of total sleep time in 24-hour polysomnography. PRS analysis identified significant associations between IH and sleep traits such as eveningness and daytime napping, with notably large effect sizes. Importantly, MR analysis further suggested a causal role for eveningness in IH pathogenesis. These findings highlight distinct genetic subtypes within IH and reinforce the relevance of circadian misalignment in its pathophysiology.