M. Shin, E. Tonini, J. S. Carpenter, B. L. Mitchell, B. Couvy-Duchesne, S. H. Park, R. Parker, E. M. Byrne, S. Hockey, A. Treneman, T. To, N. Bhattacharya, A. Shim, E. M. Scott, N. A. Gillespie, N. G. Martin, J. Scott, S. Medland, J. Crouse, I. B. Hickie
Background: Circadian rhythm disturbances (CRDs) are transdiagnostic risk factors across mental disorders. We examined CRD-associated phenotypes, their heritability, and clinical, functional and genetic correlates, cross-sectionally, in young adults from the Brisbane Longitudinal Twin Study. Methods: We assessed 2,773 community-based participants (25.6{+/-}4.1 years; 57.8% female). Six CRD-associated phenotypes (seasonality, hypersomnia, social jetlag, delayed sleep, evening preference, sleep inertia) were included. Associations of CRD-load (number of CRDs) with clinical and functional outcomes, and with polygenic risk scores (PRS) for mental health, physical health, and sleep/circadian traits, were tested using linear, negative binomial, or logistic regression, with family-clustered standard errors. Heritability of CRD-phenotypes was estimated using structural equation modelling. Results: Almost half the sample (49.2%) reported experiencing [≥]1 CRD-phenotype and 20.5% reported [≥]2. Four CRD-phenotypes (hypersomnia, social jetlag, delayed sleep, evening preference) were significantly heritable (h2=0.19-0.48). Higher CRD-load was associated with greater psychological distress ({beta}=0.96, p<0.001), more psychological symptoms (incidence rate ratio [IRR]=1.07-1.40), higher odds of a full-threshold mental disorder (OR=1.28, p<0.001), and greater functional impairment (days-in-bed, IRR=1.55; days-out-of-role, IRR=1.43). In genotyped participants (n=2,301), CRD-load was associated with mood-disorder PRS (major depression [OR=1.14, p=0.003] and bipolar disorder [OR=1.14, p=0.002]) but not PRSs for schizophrenia, attention deficit hyperactivity disorder, autism, neuroticism or anxiety. Trends towards increased PRSs for metabolic traits did not survive Bonferroni-correction. Conclusions: CRD-load was associated with clinical symptoms, functional impairment, and mood-related genetic liability. These findings support further evaluation of brief assessments of circadian-associated phenotypes for identifying young people at greater clinical burden and informing circadian-focused prevention and early intervention strategies.