Background: Sleep problems are common in psychotic disorders and are associated with worse quality of life and disease prognosis. Genome-wide association studies (GWAS) have revealed genetic influences for schizophrenia and sleep, but polygenic scores (PGSs) for sleep traits have not been evaluated systematically in patients with psychotic disorders.

Methods: This study investigated the associations between PGSs for sleep traits (insomnia, PGS_INS; sleep duration, PGS_SD; short sleep duration, PGS_SS; long sleep duration; PGS_LS), diurnal preference (eveningness, PGS_ME), and schizophrenia (PGS_SZ) with clinical features of psychotic disorders in the Finnish SUPER study comprising 8,232 patients with psychotic disorders. The measures included self-reported sleep and well-being, cognitive assessments, clozapine use, and functional outcomes. Using FinnGen data of 356,077 individuals, we analyzed the distributions of PGSs in psychotic and bipolar disorders and the general population.

Results: PGS_INS associated with more sleep problems and worse well-being (e.g. worse health-related quality of life [β = -0.07, CI = -0.09, -0.05, p < .001]). High PGS_SZ is associated with better sleep quality, worse clinical outcomes, and performance in cognitive tests (e.g. more errors in paired-associated learning [β = 0.07, CI = 0.04, 0.09, p < .001]). PGS_INS was higher in affective psychotic and bipolar disorders, while PGS_SD and PGS_ME were higher in schizophrenia as compared with individuals with no psychiatric disorders.

Conclusion: Genetic risks for sleep and diurnal preference vary between non-affective psychosis, affective psychosis, and the general population. The findings in this study emphasize the heterogeneity in genetic etiology of the objective features of disease severity and the more subjective measures related to well-being and self-reported measures of sleep.