Heterogeneity of psychiatric symptom profiles in sleep clinic outpatients: A cross-cutting dimensional approach.

Introduction: Patients accessing sleep services present with a wide range of psychiatric symptoms, but these symptoms are not well characterised. This study aims to (a) characterise psychiatric symptom profiles in patients attending an outpatient sleep clinic, (b) identify heterogenous subgroups, and (c) explore differences in pre-treatment characteristics among subgroups. Method(s): Data was collected at a university-based multidisciplinary outpatient sleep clinic via opt-out consent. Before treatment, patients completed the Cross-Cutting Symptom Measure (CCSM), which uses a dimensional approach to measure 13 mental health domains (e.g., depression, anger, mania, anxiety, psychosis, memory) and provides a transdiagonistic symptom profile. Other measures included PROMIS Sleep Disturbance and Daytime Functioning, Insomnia Severity Index, Morningness-Eveningness Questionnaire Reduced, demographics, and polysomnography (PSG; subgroup). Result(s): 352 participants (55.2% female; age M +/- SD = 43.52 +/- 16.37) who completed the CCSM were included. Latent class analysis revealed 3 subgroups with distinctive psychiatric profiles: Affective-Disturbance subgroup (n = 199, 56.5%) reported moderate affective symptoms such as depression, anxiety, and anger; Psychopathology subgroup (n = 76, 21.6%) not only reported severe affective symptoms, but also expressed more severe psychopathology such as suicidality, psychosis, and dissociation; Low-Symptom subgroup (n = 77, 21.9%) reported mild affective symptoms and few other symptoms. Psychopathology subgroup is the youngest, showed the strongest in evening chronotype, and reported the highest sleep disturbance and daytime impairment (p-values <0.01). Discussion(s): Patients attending a sleep service presented with a diverse range of psychiatric symptoms, and symptom profiles were heterogenous. Subgroups identified in this study could inform subsequent treatment planning. In our sample, affective disturbance is a predominant feature