Abstract Introduction Sleep difficulties coincide with psychosis-spectrum (PS) symptoms among individuals at clinical high risk for psychosis (CHR) and those experiencing their first episode of psychosis (FEP). Limited research has demonstrated time-linked associations between sleep and PS symptoms in adolescents at CHR or with FEP. The present study explored the associations between PS symptoms and sleep behaviors among psychiatrically acute adolescents. Methods Psychiatrically hospitalized adolescents (n=1164; 14.95±1.75 years; 64.9% female) reported on 1) PS symptoms (PRIME Screen- Revised), 2) sleep quality (PROMIS sleep disturbance and sleep-related impairment scales), 3) depression, anxiety, and psychological stress (PROMIS scales). A subsample completed a clinician-rated PS assessment followed by 2 weeks of objective sleep monitoring (wearable Actigpatch on dominant triceps; scored via Sadeh) and twice-daily diaries of sleep and PS symptoms. Linear regressions and correlations explored associations among PS symptoms and sleep characteristics, including self-reported sleep quality and actigraphy-derived sleep parameters (i.e. total sleep time/duration TST, sleep efficiency SE, time awake after sleep onset WASO, and the sleep regularity index SRI). Results In the inpatient sample, greater PS symptoms were significantly associated with worse sleep disturbances (beta=.12, t=4.16, p.001) and, separately, greater sleep impairment (beta=.12,t=4.05, p.001), beyond the significant effects of stress, depressive symptoms, anxiety, and sex. In the actigraphy sample, more diary-reported hallucinations were significantly correlated with shorter TST (r=-.62, p=.013) and less SE (r=-.52, p=.045), but not with WASO or SRI (p’s .05). Paranoia reports were not significantly associated with any of the actigraphy-derived sleep parameters (p’s .05). Greater clinician-rated positive symptom totals were associated with decreased sleep regularity (r=-.54, p=.032). Associations between clinician-rated symptoms and other sleep parameters were non-significant. Conclusion These findings indicate that PS symptoms are uniquely associated with sleep quality, beyond the effects of concurrent mental health symptoms. Discussion will include elaboration on associations between specific PS symptoms and sleep parameters, and implications of time-linked patterns observed. Further research exploring time-varying associations between PS symptoms and actigraphy-derived sleep behaviors, like those found in this study, will help elucidate underlying mechanisms linking these phenomena and inform the development of sleep-based interventions for these youth. Support (if any) P20GM139743 (PI: Carskadon)
Thompson et al. (2026) studied this question.