S6. SLEEP ENDOPHENOTYPES OF SCHIZOPHRENIA: A HIGH-DENSITY EEG STUDY IN DRUG-NAïVE, FIRST EPISODE PSYCHOSIS PATIENTS

Abstract

BackgroundSlow waves, the hallmark of the deep nonrapid eye movement sleep electroencephalogram (EEG), are critical for restorative sleep and brain plasticity. They arise from the synchronous depolarization and hyperpolarization of millions of cortical neurons and their proper generation and propagation relies upon the integrity of widespread cortico-thalamic networks. Slow wave abnormalities have been reported in patient with Schizophrenia, although with partially contradictory results, probably related to antipsychotic and sedative medications. Recently, their presence and delineation, have been convincingly shown in first-episode psychosis patients (FEP). However, clear evidence of this biomarker at the onset of the disease, prior to any psychopharmacological intervention, remains limited. Moreover, no attempt has been made to elucidate the prognostic meaning of this finding.MethodsWe collected whole night sleep high–density electroencephalography recordings (64-channel BrainAmp, Brain Products GmbH, Gilching, Germany) in 20 drug-naive FEP patients and 20 healthy control subjects (HC). Several clinical psychometric scales as well as neurocognitive tests were administered to all subjects in order to better define psychopathological status and vulnerability. EEG slow wave activity (SWA, spectral power between 1 and 4 Hz) and several slow wave parameters were computed at each electrode location, including density and amplitude, at each electrode location. Along with a group analysis between FEP and HC, a subgroup analysis was also computed between patients who showed a progression of symptoms to full-blown Schizophrenia (SCZ, n = 10) over the next 12-month follow-up and those who did not (OTH, n = 10).ResultsSleep macro-architecture was globally preserved in FEP patients. SWA (1–4 Hz) was lower in FEP compared to HC but this difference didn’t reach statistical significance. Slow wave density was decreased in FEP compared to HC, with a significance that survived multiple comparison correction over a large fronto-central cluster. Mean amplitude was preserved. At the subgroup analysis, these results were largely driven by the subgroup of patients with a confirmed diagnosis of SCZ at a 12-month follow-up. Indeed, no difference could be found between OTH and HC, while a strong significance was still evident between SCZ and HC.DiscussionOur data confirm previous findings on reduced slow wave density in FEP, and expand them to acute subjects, before any treatment is prescribed. This is in line with available data on diffuse abnormalities of cortico-cortical and cortico-thalamic networks in these patients. Interestingly, our data also offer preliminary evidence that this deficit is specific for SCZ, as it appears to differentiate patients who developed SCZ from those with other diagnoses at follow-up. Given the traveling properties of slow waves, future research should establish their potential as markers of connectivity in SCZ.