Reduced slow wave density is associated with worse positive symptoms in clinical high risk: an objective readout of symptom severity for early treatment interventions?

Abstract

Individuals at clinical high risk for psychosis (CHR) present subsyndromal psychotic symptoms that can escalate and lead to the transition to a diagnosable psychotic disorder. Identifying biological parameters that are sensitive to these symptoms can therefore help objectively assess their severity and guide early interventions in CHR. Reduced slow wave oscillations (~1 Hz) during non-rapid eye movement sleep were recently observed in first-episode psychosis patients and were linked to the intensity of their positive symptoms. Here, we collected overnight high-density EEG recordings from 37 CHR and 32 healthy control (HC) subjects and compared slow wave (SW) activity and other SW parameters (i.e., density and negative peak amplitude) between groups. We also assessed the relationships between clinical symptoms and SW parameters in CHR. While comparisons between HC and the entire CHR group showed no SW differences, CHR individuals with higher positive symptom severity (N=18) demonstrated a reduction in SW density in an EEG cluster involving bilateral prefrontal, parietal, and right occipital regions compared to matched HC individuals. Furthermore, we observed a negative correlation between SW density and positive symptoms across CHR individuals, suggesting a potential target for early treatment interventions.

1. Introduction

Sleep disturbances are a common feature of psychotic disorders, and subjective sleep disturbances have been consistently reported in individuals at clinical high-risk for psychosis (CHR) (Lederman et al., 2017; Lunsford-Avery et al., 2015; Zaks et al., 2022; Zanini et al., 2015). Previous studies have investigated sleep architecture in this population (Lunsford-Avery et al., 2015; Mayeli et al., 2021; Purple et al., 2020; Zanini et al., 2015), revealing lower sleep efficiency (Lunsford-Avery et al., 2015) and higher sleep latency (Zanini et al., 2015) and wakefulness after sleep onset (Mayeli et al., 2021; Zanini et al., 2015) in CHR compared to healthy control (HC) subjects. However, as highlighted by a recent meta-analysis from our group, a gap remains in the literature regarding specific investigations of slow wave abnormalities and spindle characteristics in at-risk individuals (Bagautdinova et al., 2023). In a recently published study, we examined sleep spindle features in CHR compared to HC subjects (Mayeli et al., 2022c) and established reduced spindle duration in CHR individuals in centro-parietal and prefrontal regions, indicating alterations in sleep neurophysiology within this population. Of note, spindle deficits have been reported in chronic, early course, and first-episode psychosis (FEP) patients, thus suggesting that they may represent a trait neurophysiological biomarker for psychotic disorders (Lüthi, 2014).

Beside sleep spindles, slow waves are the other main oscillatory brain activity occurring during non-rapid eye movement (NREM) sleep, and contribute to various neural processes (Dijk, 2009). Slow waves are characterized by ~1 Hz EEG oscillations and reflect the synchronized interactions between cortical neurons, manifesting as alternating up and down states of neuronal depolarization and hyperpolarization respectively (Adamantidis et al., 2019). Our group has recently shown a reduction in slow wave density during sleep in FEP patients in frontal-prefrontal regions, and this reduction was associated with the severity of the positive symptoms experienced by these individuals (Kaskie et al., 2019). Intriguingly, slow wave density does not appear to differ significantly in individuals with chronic schizophrenia (Ferrarelli et al., 2010; Kozhemiako et al., 2022), suggesting that this sleep abnormality may be particularly relevant in the early stages of psychosis and could serve as a potential state biomarker for these disorders.

In this study, our primary objective was to examine the presence of slow wave abnormalities during sleep in CHR individuals. We employed high-density electroencephalography (hd-EEG) recordings to comprehensively investigate various parameters related to slow wave activity, including activity, density, and amplitude. Additionally, we aimed at investigating the relationship between slow wave abnormalities and subthreshold psychotic experiences in CHR subjects.

2. Methods

2.1. Recruitment, Eligibility Criteria, and Assessments

CHR participants were recruited through UPMC clinical settings, referrals from other clinicians, and outside sources, while HC subjects were recruited from the local community using physical and online advertisements. Written informed consent was obtained from all study participants prior to any assessments. The University of Pittsburgh Institutional Review Board approved the study, which was conducted at Western Psychiatric Hospital (WPH) in Pittsburgh.

Study participants underwent a comprehensive screening, including the structured clinical interview for DSM disorders (SCID) and the Structured Interview for Prodromal Symptoms (SIPS). The SIPS interview is employed to rate subthreshold-intensity symptoms across four major dimensions on the Scale of Prodromal Symptoms (SOPS): positive (5 items), negative (6 items), disorganized (4 items), and general (4 items) symptoms (Miller et al., 2003). Eligibility criteria in both groups included: 1) ages between 12 and 35 years old; 2) no lifetime history of neurological disorder or head injury resulting in loss of consciousness for more than 1 minute, 3) being able to speak English fluently enough to participate study procedures; 4) being able to travel to WPH; 5) no pregnancy, 6) no history of alcohol or drug dependence in the past year. Additional exclusion criteria for HC individuals were as follows: 1) a psychiatric disorder diagnosis, assessed with the SCID; 2) CHR state; 3) first-degree relatives with a diagnosed psychotic disorder. Additional eligibility criteria for the CHR group were a subsyndromal score on one of the five SOPS positive symptom items, along with a worsening trajectory over the previous year.

2.2. Positive Symptom Assessment

In addition to SOPS positive symptoms (SOPS-PS) assessment, CHR participants completed the scale for the assessment of the positive symptoms (SAPS) (Andreasen, 1984). SAPS provides a more comprehensive evaluation of positive symptoms and has been utilized in both CHR and schizophrenia patients (Fulford et al., 2013; Ruiz-Castañeda et al., 2022). SAPS includes 34 items and assesses four subscales related to positive symptoms: hallucinations, delusions, bizarre behavior, and positive formal thought disorder.

2.3. Sleep hd-EEG recordings

Two whole night sleep EEG recordings were acquired for all participants at the University of Pittsburgh Sleep and Behavioral Neuroscience Center. Due to the potential first-night effect on sleep EEG parameters (Mayeli et al., 2022b) , only data from the second night was analyzed for the present study. We utilized a 128-channel high density (hd)-EEG system (Electrical Geodesics INC, EGI, Eugene, Oregon) to collect overnight sleep EEG recordings at a sampling rate of 250 Hz. Conductive gel was applied to each electrode to obtain impedance values below 100 KΩ for all channels. Experienced sleep technicians used American Academy of Sleep Medicine (AASM) criteria (Silber et al., 2007) to score overnight sleep EEG recordings.

2.4. Sleep hd-EEG data processing

Sleep hd-EEG data was analyzed using MATLAB (The MathWorks Inc., Natick, MA) with the following steps. The EEG data was initially band-passed filtered between 0.5 Hz and 40 Hz, and downsampled from 250 Hz to 128 Hz. Then, after excluding the channels with a significant difference between the highest and lowest values the EEG data was re-referenced to the average of all electrodes and segmented into 6-s epochs for computation of sleep EEG power spectra. To transform the sleep EEG time series into the frequency domain, Welch’s modified periodogram method was applied using two second Hamming windows (with 50% overlap). A semi-automatic artifact rejection procedure was employed to remove channels and epochs with high-frequency noise or with other artifacts, such as muscle activity. The artifact rejection thresholds for low (1–4 Hz) and high (20–30 Hz) frequency ranges were automatically calculated for each channel individually. The higher threshold was utilized to account for muscle artifacts since muscle activity typically falls within that frequency band, while the lower threshold was used to detect low-frequency drift caused by poor channel contact and sweating. Epochs with significant artifacts were also rejected based on a 36-sec sliding window, which was set to be three times higher than the low-frequency power and six times higher than the high-frequency power of that window. For each channel, the sleep EEG spectral power in the low and high frequency bands was plotted and all 6-s NREM epochs were visually inspected. The EEG channels wherein artifacts affected the majority (> 50%) of the recording were excluded.

Topographic power analyses were performed within the slow wave range (0.5–4 Hz) to examine possible slow wave activity differences between CHR and HC groups. To further analyze and characterize slow wave parameters, a customized algorithm was applied for the automatic detection of sleep slow waves (Ferrarelli et al., 2010; Kaskie et al., 2019). Specifically, the sleep EEG signal was re-referenced to the average of the two mastoid electrodes and band-pass filtered at 0.5–4.0 Hz. Slow waves were identified as negative deflections occurring between two zero crossings and were selected based on their period (i.e., only waves with consecutive zero crossings spanning 0.25 to 1.0 seconds, detected in artifact-free NREM epochs, were considered as slow waves). Two slow wave parameters, negative peak amplitude (i.e., the amplitude of the peak occurring after the first zero crossing) and density (i.e., the number of slow waves detected per minute of NREM sleep), were computed and compared across CHR and HC participants. The negative peak amplitude provides a direct insight into the intensity of slow wave synchronization (Esser et al., 2007), while density captures the overall incidence of slow waves regardless of their amplitude (Kaskie et al., 2019).

As an exploratory analysis, we conducted SW-spindle coupling analysis as performed in previous studies in patients with schizophrenia (Demanuele et al., 2017; Kozhemiako et al., 2022). Specifically, we identified the SWs with amplitudes higher than 40 μV and determined the phase at the spindle peak to quantify the SW-spindle coupling. For more details of the spindle detection, please refer to (Mayeli et al., 2022c).

2.5. Statistical analyses

We conducted an unpaired two-sided Student’s t-test and a chi-squared test to evaluate possible age and sex differences between the two groups, respectively. Unpaired t-tests in MATLAB were used to investigate variations in individual sleep slow wave features (i.e., slow wave activity, negative peak amplitude, and density) between the CHR and HC groups. To account for multiple comparisons due to the number of channels, we performed a clustering analysis on the t-map of each slow wave feature using a Gaussian mixture model (GMM) (Fraley and Raftery, 2002). We set the number of clusters for each t-map to be the value that achieved the minimum adjusted Bayesian information criterion (BIC). These analyses were performed in R using the package mclust (Scrucca et al., 2016). A Bonferroni correction was applied to account for the topographic features (i.e., the scalp regions) that were part of the significant cluster identified with the Gaussian mixture model. The Cohen’s D effect size was also computed for the region(s) showing the largest difference between CHR and HC participants.

We utilized the z-scores of SAPS and SOPS-PS to stratify CHR subjects into subgroups based on their positive symptom scores. We calculated the z-scores for SAPS and SOPS-PS separately, and then summed those scores. Subjects with a sum greater than zero were classified into the high positive symptoms’ subgroup, while those with a sum less than zero were assigned to the low positive symptoms’ subgroup. Age and sex-matched HC individuals were then compared with each of the CHR subgroups. Unpaired t-tests were subsequently repeated to evaluate potential differences in slow wave features between the CHR subgroups and their respective matching HC subgroups. Furthermore, we conducted Pearson correlation analyses to explore the relationships between the summed z-score of SAPS and SOPS-PS, as well as individual scores of SAPS and SOPS-PS, with slow wave density across all CHR participants.

3. Results

3.1. Demographic and clinical characteristics

Thirty-seven CHR individuals (14 female) and thirty-two age-matched HC (13 female) subjects were included in this study. Table 1 provides an overview of the demographic and clinical measures for each study group. There were no differences in age or sex between the CHR and HC groups.

Table 1:

Demographic and Clinical variables of study groups.

Clinical Measures	Healthy Controls	Clinical High Risk	P-Value
Number of subjects	N = 32	N = 37
Sex (# female)	N = 18	N = 23	0.6181
Age in years (average ± SD)	21.0 ± 4.9 (14 – 33)	21.0 ± 4.1 (13 – 30)	0.9852
SOPS positive symptoms (average ± SD)	NA	11.9 ± 3.0
SOPS negative symptoms (average ± SD)	NA	11.5 ± 5.1
SOPS disorganized symptoms (average ± SD)	NA	5.1 ± 2.1
SOPS general symptoms (average ± SD)	NA	7.5 ± 3.0
SAPS	NA	28.7 ± 7.3
SANS	NA	5.9 ± 4.3

SD represents the standard deviation.

SAPS: the scale for the Assessment of Positive Symptoms

SOPS: the Scale of Prodromal Symptoms-Positive Subscale

¹Pearson Chi-Square 2-sided p-value.

²paired t-test 2-sided p-value.

3.2. Sleep architecture parameters and sleep EEG power analysis

Compared to HC subjects, CHR individuals had a significantly higher sleep onset latency: p = 0.031, t-stat = 2.200 as well as a lower sleep efficiency: p = 0.017, t-stat = −2.448. CHR also had a higher percentage of wake after sleep onset (WASO) that showed trend level significance (p = 0.085, t-stat = 1.743). No differences in any other sleep architecture parameters, including total sleep time, percentage of time spent in NREM (i.e., N1, N2, and N3), and rapid eye movement (REM) sleep stages were found between the two study groups (Table 2).

Table 2:

Sleep architecture variables for healthy control (HC) and clinical high risk (CHR) for psychosis groups.

Sleep Architecture Parameter	CHR	HC	CHR vs HC
	Mean (SD)	Mean (SD)	Uncorrected P-value	t-stats	df
Total sleep time (min)	433.335 (74.018)	454.297 (47.370)	0.179	−1.357	67
Sleep latency (min)	27.127 (21.863)	17.131 (13.730)	0.031	2.201	67
Sleep efficiency	85.9% (10.0%)	90.9% (5.7%)	0.017	−2.448	67
NREM N1 (%)	6.8% (5.5%)	6.6% (3.3%)	0.804	0.249	67
NREM N2 (%)	49% (8.8%)	48.7% (5.9%)	0.866	0.170	67
NREM N3 (%)	23.6% (10.5%)	23.3% (7.6%)	0.871	0.164	67
REM (%)	20.5% (6.5%)	21.5% (4.7%)	0.495	−0.685	67
WASO (%)	8.5% (8.3%)	5.6% (4.9%)	0.085	1.743	67

NREM: non-rapid eye movement sleep

REM: rapid eye movement sleep

WASO: Wakefulness after sleep onset

SD: represent the standard deviation

Figure S1 shows the mean NREM sleep EEG power spectra for the healthy control (HC, blue line) and clinical high risk (CHR, red line) groups. The shaded areas represent 95% confidence intervals for the mean power at each frequency band per group, pooled across good EEG channels. While no between-group power differences were observed in the delta, theta, alpha or sigma bands, significantly higher beta and gamma power was seen in CHR versus HC (uncorrected for multiple comparisons). Furthermore, the topographical comparisons of power across EEG frequency bands revealed no significant between-group differences in delta, theta, alpha, or sigma power. However, higher beta and gamma power was observed in frontal and posterior channels in CHR compared to HC individuals (Figure S2).

3.3. Slow wave measures in CHR and HC individuals

Topographic analyses showed no differences in slow wave activity between CHR and HC individuals (Figure S3). Also, no significant differences were found in slow wave negative peak amplitude and density when comparing the two study groups, after cluster analysis (Figure S4). Building upon our recent findings about the correlation between reduced slow wave density and higher positive symptoms in FEP patients (Kaskie et al., 2019), we stratified the CHR into high and low positive symptoms subgroups, resulting in 18 subjects with higher positive symptoms and 19 subjects with lower positive symptoms. No differences in SWA were observed when comparing subgroups of CHR based on positive symptoms with matched HC individuals (Figure S5). In contrast, CHR individuals with high positive symptoms showed lower slow wave density compared to matched HC subjects (Figure 1A) in a large cluster involving bilateral prefrontal, parietal, and right occipital regions (t = −2.25; p = 0.03; Cohen’s d = −0.75), while no significant differences in slow wave density were observed between CHR individuals with low positive symptoms and matched HC individuals (Figure 1B).

Figure 1: Clinical high risk for psychosis (CHR) and healthy control (HC) differences in slow wave (SW) density.

A). A reduction in slow wave density was observed in CHR individuals with high positive symptoms and matched HC in a large cluster involving bilateral prefrontal, parietal, and right occipital regions. White dots indicate the channels that contributed to the significant cluster (t= −2.25; p=0.03, after clustering analysis). (B). No significant differences in slow wave density were observed between CHR with low positive symptoms and HC subjects.

To ensure that the observed between-group differences in that cluster were not influenced by the subset of HC participants who are matched to the CHR subgroups, we compared each CHR subgroup with all 32 HC individuals. This analysis confirmed the CHR subgroup findings and showed a reduction in slow wave density in CHR with high, but not low, PS relative to HC (Figure S6).

We further compared SW density between CHR individuals with high PS and matched HC during cycle 1 sleep. Though topographical maps showed a similar pattern, no significant between-group differences in SW density emerged at any specific cluster of channels after correcting for multiple comparisons (Figure S7).

We also compared SW parameters between CHR individuals with high and low PS and found no differences in any of those parameters (Figure S8) after multiple cluster correction.

Furthermore, an examination of the slow wave phase at the peak of sleep spindle amplitude revealed no significant differences between CHR and HC groups, even when assessing the more symptomatic CHR individuals (Figure S9).

3.4. Relationship between slow-wave density and positive symptoms in CHR

We established a robust, significant negative correlation between the average slow wave density from that cluster and the sum z-score of SAPS and SOPS-PS (R = 0.47, p = 0.003; Figure 2A). Additional correlation analyses revealed a strong negative association between slow wave density and SAPS (R = 0.48, p = 0.003; Figure 2B), while a trend towards a significant negative correlation was observed with SOPS-PS (R = 0.29, p = 0.083; Figure S10). Furthermore, the associations between positive symptoms and SW density were non-significant when we examined CHR individuals with high and low PS separately, likely due to the lack of PS variability in each subgroup (Figure S11).

Figure 2: Negative correlations between slow wave density and positive symptoms in clinical high risk for psychosis (CHR) individuals.

A significant negative correlation was established between reduced slow wave density and the sum z-score of the assessment of the positive symptoms (SAPS) and SOPS positive symptoms (R = −0.47, p = 0.003) (A) as well as the SAPS total score (R = −0.48, p = 0.003) (B). Individuals with high positive symptoms in CHR are represented by red dots, whereas those with low positive symptoms are depicted with blue dots.

4. Discussion

By performing sleep high-density electroencephalography recordings, in this study, we investigated differences in slow wave features, including activity, negative peak amplitude, and density, in CHR individuals vs. HC subjects. After stratifying CHR based on clinical symptom severity, we established that CHR individuals with higher positive symptoms, but not those with lower levels of symptoms, had reduced slow wave density, particularly in bilateral prefrontal, parietal, and right occipital regions relative to HC subjects. Furthermore, we established a negative correlation between reduced slow wave density and the severity of positive symptoms for the CHR.

In this study, we found altered sleep architecture parameters in CHR relative to HC groups, including higher sleep latency, lower sleep efficiency, and a trend-like increase in WASO. hese findings are in line with recent work from our group showing sleep architecture abnormalities in CHR compared to HC individuals (Mayeli et al., 2021; Mayeli et al., 2022c). Furthermore, the power analyses showed no difference in any frequency bands except in high beta and gamma, consistent with a previous sleep EEG study where we found higher gamma in CHR vs HC subjects (Mayeli et al., 2021).

The decrease in slow wave density that we established here in CHR with higher positive symptoms was similary to the one observed in our recent study conducted on individuals with FEP (Kaskie et al., 2019). The reduction in density was specific for both CHR and FEP individuals, given that other slow wave parameters, including amplitude and power did not differ between these two clinical populations and HC groups. Furthermore, the finding of specific slow wave density deficits in both CHR and FEP, particularly in prefrontal and parietal regions, suggests a similar pattern of impairment in these two clinical populations, and is consistent with alterations in prefronto-parietal regions, including prefrontal cortex, parietal lobe, and posterior cingulate cortex, that have been reported by several fMRI studies in patients with psychosis and schizophrenia (Fryer et al., 2013; Gisselgård et al., 2018; Shim et al., 2010). In addition to prefrontal-parietla regions, in CHR individuals, the reduction in slow wave density also involved more posterior, occipital regions and was especially prominent on the right side. Of note, it has been recently shown slow-wave density tends to be higher in more posterior regions, especially on the right hemisphere, in younger individuals (Castelnovo et al., 2023). Given that CHR are a younger population than FEP, the implication of more posterior, parietal and occipital regions especially on the right side in CHR vs. age matched HC may reflect this developmental characteristic of sleep slow waves. Another possible explanation is the higher resolution of the CHR (i.e., N=128 channels) vs. the FEP (i.e., N=64 channels) studies.

Slow wave findings in patients with psychosis and schizophrenia have been mixed. While some sleep studies have reported a deficit in slow wave activity and density among individuals with schizophrenia (Hiatt et al., 1985; Keshavan et al., 1998; Sekimoto et al., 2011), especially those who were actuely ill and/or drug-naïve (Hiatt et al., 1985; Keshavan et al., 1998), several other sleep studies failed to establish any significant slow wave abnormalities in chronic schizophrenia patients when compared to HC (Genzel et al., 2015; Göder et al., 2015; Manoach et al., 2010). Notably, more recent hd-EEG sleep studies conducted on these patients did not find significant differences in slow wave density (Ferrarelli et al., 2010; Kozhemiako et al., 2022). This lack of differentiation could potentially be attributed to the stage of the illness and/or the impact of medication, which is commonly used to manage and reduce positive symptoms in patients with schizophrenia. Indeed, while negative symptoms and cognitive deficits pose challenges in terms of treatment response, medications have shown effectiveness in controlling and alleviating positive symptoms (Mayeli et al., 2022a; Millan et al., 2014). Consistent with this assumption, we observed a significant negative correlation between slow wave density and the severity of positive symptoms, indicating that lower slow wave density was linked to higher symptom severity in CHR individuals. Therefore, these findings support the notion that deficits in slow wave density may serve as a state-related biomarker, reflecting the severity of positive symptoms in individuals at risk for or in the early stages of psychotic disorders.

The molecular mechanisms underlying sleep slow wavrs have yet to be fully established. However, serotonin, glutamate, and dopamine neurotransmission have been implicated in sleep and slow wave regulation (Monti and Jantos, 2008). Imbalances in those neurotransmitters can disrupt sleep and contribute to psychosis susceptibility, particularly relevant to individuals at CHR. Hallucinogens such as NMDA antagonist PCP and the 5-HT2A serotonin agonist 2,5-dimethoxy-4-iodoamphetamine (DOI) induce psychosis in HC individuals (Javitt et al., 2012) and reduce slow wave power in rats’ prefrontal cortex (Celada et al., 2008). In contrast, Clozapine, one of the most effective antipsychotic compounds in patients with schizophrenia, can counter DOI’s effect (Celada et al., 2008); nonetheless, the ability of clozapine and other antipsychotic medicines to modulate slow wave density in these patients has yet to be established. Animal studies have shown that sleep deprivation reduces sensorimotor gating (PPI) (Frau et al., 2008), which is impaired in patients with schizophrenia (Braff, 2010), whereas D2 antagonists and post-recovery sleep reverse such deficits (Frau et al., 2008). Enhanced D1/D2 receptor stimulation, which is associated with higher proneness to psychosis, can disrupt sleep, while dopamine-blocking antipsychotics reduced psychotic symptoms although, again, the impact on sleep slow wave density has been less investigated (Donati et al., 2020). Further research, manipulating dopamine and serotonin, is therefore needed to understand their implications in reduced slow wave density in CHR individuals.

Conducting hd-EEG assessments, as performed in this study, in larger populations of CHR individuals as well as in patients with early psychosis in the sleep laboratory can be a challenging undertaking. However, wearable and easy to set-up devices were recently developed that allow for the assessment of objective sleep measures in the home environment (Kwon et al., 2021; Zeller et al., 2023). While these devices still need to be validated against standard tools such as PSG, at-home sleep EEG assessment offers a more convenient and accessible approach for evaluating sleep in patients with psychotic symptoms and, more broadly, psychiatric conditions. Importantly, the sleep alterations we observed in individuals at CHR, those with FEP, and patients with schizophrenia, including reduced spindle duration across all three groups, and deficits in slow wave density in CHR and FEP, were primarily localized in frontal/prefrontal channels (Ferrarelli et al., 2007; Kozhemiako et al., 2022; Mayeli et al., 2021; Mayeli et al., 2022c). These findings suggests that portable sleep EEG devices available on the market could be effectively employed to assess these sleep abnormalities in a practical and efficient manner.

Further research is required to explore and provide answers to the remaining questions raised by the current study. First, these findings need to be replicated in larger cohorts of CHR individuals with sensitivity analysis based on positive symptom severity. The CHR population exhibits heterogeneity, characterized by variations in symptoms, functioning, and illness trajectories (Fusar-Poli et al., 2016; Mittal and Addington, 2021). This diversity presents challenges in research and clinical contexts, necessitating tailored approaches (Thompson et al., 2015). Future studies should therefore examine the sensitivity and specificity of reduced slow wave density for worse positive symptoms in larger CHR cohorts. Second, in this study symptom severity was assessed both with the SOPS positive score and with the summation of the SAPS and SOPS positive symptom z-scores and we found that the latter had a stronger correlation with slow wave density deficits. Future work is needed to determine whether a more detailed assessment of psychotic assessment, as in the SAPS, the positive and negative syndrome scale (PANSS), and the brief psychiatric rating scale (BPRS) is necessary to identify these slow wave deficits in CHR (De Pablo et al., 2021). Third, while our findings indicate slow wave density as a potential state maker for psychosis, longitudinal studies are needed to first examine whether relationship between slow wave density and negative symptoms will be confirmed in follow up assessments and whether slow wave density can predict the trajectory of the disorder.

Altogether, findings from this study provides compelling evidence of deficits in slow wave density, which were most prominent in bilateral prefrontal, parietal, and occipital regions, among individuals at CHR with higher positive symptom severity relative to HC subjects. Building on these significant findings, future research efforts will contribute to a deeper understanding of the critical role that slow wave and related sleep deficits play in the pathophysiology and symptom severity during the early phases of psychosis. Moreover, this line of investigation may pave the way for the development of more effective early treatment interventions, both for at-risk individuals and patients with psychosis, based on sleep neurophysiology. For example, in addition to monitoring sleep abnormalities in individuals at risk and patients with psychotic disorders through at-home tracking, recent technological advancements have been developed and showed viability of enhancing sleep slow wave. These include closed loop auditory stimulation using portable devices during sleep (Garcia-Molina et al., 2018; Zeller et al., 2023) and transcranial electrical and magnetic stimulation during wakefulness (Grimaldi et al., 2020; Wilckens et al., 2018), rendering these defecits an accessable target for non-invasive and drug-free interventions.

Highlights:

• Reduced SW density reported in patients with early-stage psychotic disorders

• Here, we compared SW parameters between CHR and HC individuals

• CHR with higher positive symptoms exhibited lower sleep SW density vs. HC

• In CHR, slow wave density was inversely correlated with positive symptoms

• Low SW density in CHR implies a potential target for early treatment interventions