Mediators and moderators of outcome from the Transdiagnostic Sleep and Circadian Intervention for adults with severe mental illness in a community setting

Abstract

This study explores mediators and moderators of outcome from the Transdiagnostic Sleep and Circadian Intervention (TranS-C). Participants were 121 adults (age mean [SD] = 45.45 [13.25] years; 52.06% female), with severe mental illness and sleep and circadian dysfunction. They were randomized to receive TranS-C and usual care, or usual care followed by delayed TranS-C. Putative mediators were the PROMIS-Sleep Disturbance Scale, PROMIS-Sleep-Related Impairment Scale, and the Sleep Health Composite. Putative moderators were symptom severity at baseline, age, housing environment, race and ethnicity. Mediation analysis indicated that changes in sleep-related impairment and sleep health, but not sleep disturbance, mediate TranS-C's effect on functional impairment (CI: [-0.28,-0.07] and [-0.18,-0.08], respectively) and general psychiatric symptoms (CI [-0.30,-0.06] and [-0.20,-0.02]) post-treatment, though mediation via sleep health was not maintained through 6-month follow up. Moderation analysis indicated that those with severe sleep-related symptoms at baseline and Black/African American individuals are particularly responsive to TranS-C, whereas older recipients had a poorer response. TranS-C reduces functional impairment and general psychiatric symptoms amongst adults with SMI by reducing sleep-related impairment and improving sleep health. Selected sleep outcomes of TranS-C are strongest among Black/African Americans and those with severe sleep problems, and weakest among older adults.