Sleep as a vulnerability marker in clinical high-risk populations. a multimodal systematic review

Sleep disturbances are increasingly recognized as clinically significant features of the psychosis risk state, yet no comprehensive synthesis has examined sleep across all assessment modalities in clinical high-risk (CHR) populations.

Following PRISMA 2020 guidelines, five major databases were searched (January 1995-March 2026) for original studies assessing sleep in CHR populations, with CHR status operationalized through validated instruments.

From 5218 citations, 26 studies were included. Sleep disturbances affected 59-80% of CHR individuals. Actigraphy predicted positive symptom worsening at 12-month follow-up. Experience sampling identified nocturnal awakenings as predictors of next-day paranoia and derealization, mediated by morning negative affect. Sleep disturbance predicted symptom severity over time, accelerated transition among converters, and was independently associated with suicidal ideation. High-density EEG polysomnography documented reduced prefrontal spindle duration, elevated non-rapid eye movement (NREM) gamma power, and reduced slow wave density - replicated across independent cohorts. A multimodal neuroimaging study identified whole-thalamus-dorsolateral prefrontal cortex (DLPFC) hypoconnectivity as a convergent thalamocortical marker.

Sleep disturbances in CHR populations span multiple neurobiologically distinct domains. Polysomnographic markers - spindle parameters and NREM gamma power - represent the most neurophysiologically specific signatures identified. Standardized sleep assessment should be integrated into CHR clinical protocols as both a prognostic tool and a modifiable therapeutic target.