The influence of novelty detection on the 40-Hz auditory steady-state response in schizophrenia: A novel hypothesis from meta-analysis

the


Introduction
Steady-state responses are oscillatory responses elicited by periodic stimuli in various sensory modalities.These responses can be detected non-invasively by electroencephalography (EEG) and magnetoencephalography (MEG).Auditory steady-state responses (ASSRs) exhibit a peak at approximately 40 Hz (Picton et al., 2003), while visual steadystate responses peak at around 10 Hz (Herrmann, 2001).Previous MEG (Ross, 2008) and positron emission tomography (Pastor et al., 2002) studies have indicated that the ASSR originates in the primary auditory cortex.However, recent research suggests that the generation of the 40-Hz ASSR involves a more extensive network that includes nonprimary cortical areas (Arnal et al., 2019;Farahani et al., 2019Farahani et al., , 2021) ) and subcortical regions (Herdman et al., 2003;Steinmann and Gutschalk, 2011;Grent-'t-Jong et al., 2021).Although the comprehensive understanding of the generation mechanisms of ASSRs remains incomplete, ASSRs are interpreted as reflections of the synchronization of endogenous oscillations (Santarelli et al., 1995;Ross et al., 2002Ross et al., , 2005)).Thus, the 40-Hz ASSR is thought to reflect the tendency of neurons to oscillate at the gamma frequency (30-80 Hz) induced by external stimulation.Gamma oscillations are linked to various cognitive processes, including sensory binding (Engel et al., 2001), attentional selection (Fries et al., 2001;Gregoriou et al., 2009), and memory (Lisman, 1999;Miller et al., 2018), which has captured the interest of researchers.Schizophrenia (ScZ), characterized by a wide range of symptoms, including delusions, hallucinations, blunted affect, anhedonia, and cognitive impairment, is associated with deficits in gamma oscillations (Uhlhaas and Singer, 2010;Hirano and Uhlhaas, 2021).In ScZ, dysfunction of parvalbumin γ-aminobutyric acid (GABA) interneurons and N-methyl-D-aspartate receptors (NMDARs), which are crucial for the generation of gamma oscillations, supports this view (Kantrowitz and Javitt, 2010;Lewis et al., 2012).
In 1999, Kwon et al. first reported that patients with ScZ exhibited reduced power and synchronization of the 40-Hz ASSR using EEG (Kwon et al., 1999).This finding has been replicated in numerous EEG and MEG studies by other research groups (Onitsuka et al., 2022).Furthermore, a meta-analysis showed a significant decrease in both spectral power and phase-locking in patients with ScZ compared to healthy controls (HCs) (Thuné et al., 2016).A recent meta-analysis, which expanded to include patients with first-episode psychosis and participants at clinically high risk for psychosis, provided further evidence for the consistency of the 40-Hz ASSR impairments (Zouaoui et al., 2023).These studies demonstrated that the 40-Hz ASSR impairments in ScZ are robust across different patient samples, stages, and paradigm designs.Therefore, the 40-Hz ASSR stands as a promising candidate biomarker that facilitates insight into the underlying pathophysiological mechanisms and promotes links to preclinical research (Grent-'t-Jong et al., 2023).However, to date, evidence regarding the association between the 40-Hz ASSR in ScZ and factors such as medication status, illness duration, and clinical symptoms has been limited and inconsistent across studies (Onitsuka et al., 2022;Grent-'t-Jong et al., 2023).Among the clinical correlates, for example, Light et al. showed associations between working memory deficits and 40-Hz intertrial phase synchronization in patients with ScZ (Light et al., 2006).Additionally, Spencer et al. reported that auditory hallucinations were associated with phase synchronization of the 40-Hz ASSR in the left hemisphere (Spencer et al., 2009).
The 40-Hz ASSR is influenced by various experimental parameters, including attention (Tiitinen et al., 1993), stimulus type (Voicikas et al., 2016), EEG/MEG analysis approach (Muthukumaraswamy and Singh, 2013), and stimulus characteristics such as stimulus intensity and carrier frequency (Ross et al., 2000;Picton et al., 2003).For example, selective attention typically increases the 40-Hz ASSR in healthy individuals (Tiitinen et al., 1993;Ross et al., 2004), unlike in patients with ScZ (Hamm et al., 2015;Coffman et al., 2022).Studies suggest that click trains produce larger (Voicikas et al., 2016) and more consistent responses (McFadden et al., 2014) than amplitude-modulated tones in healthy individuals.In patients with ScZ, clicks were reported to have a larger effect size for 40-Hz ASSR reduction than amplitude-modulated tones, resulting in improved discrimination between patients and controls (Griskova-Bulanova et al., 2018).In a study comparing the sensitivity of EEG and MEG in measuring gamma oscillations in healthy individuals, MEG demonstrated improved detectability of highfrequency activity (Muthukumaraswamy and Singh, 2013).Furthermore, MEG-based source reconstruction significantly enhanced the signal-to-noise ratio for estimating the 40-Hz ASSR (Tan et al., 2015).EEG/MEG source modeling approaches for measuring the 40-Hz ASSR using left and right superior temporal gyrus dipoles produced more valid measurements than EEG sensor measurements (Edgar et al., 2017).These source models were also effective in discriminating between HCs and patients with ScZ (Edgar et al., 2018).
On the other hand, the effects of stimulus duration and inter-stimulus interval (ISI) on the 40-Hz ASSR remain largely unexplored.Hamm et al. examined the influence of stimulus duration on the 40-Hz ASSR power in patients with ScZ using two stimulus durations of 500 and 1500 ms (Hamm et al., 2015).Specifically, when longer stimuli were presented, HC exhibited a reduction in the 40-Hz ASSR power, while patients with ScZ did not.Consequently, the magnitude of the 40-Hz ASSR reduction in ScZ was significantly more pronounced with a stimulus duration of 500 ms compared to 1500 ms (Hamm et al., 2015).In addition, Choi et al. recently demonstrated the effects of ISI on the 40-Hz ASSR in patients with ScZ (Choi et al., 2023).They investigated the effects of three different ISIs (500, 2000, and 3500 ms) on the 40-Hz ASSR in patients with ScZ and HCs.Their findings revealed a decrease in total power, and inter-trial coherence decreased in HCs as the ISI increased.Conversely, patients with ScZ exhibited minimal ISI effects, leading to no significant differences at longer ISIs.These results highlight the important role of stimulus duration and ISI in modulating the 40-Hz ASSR.
The two previous meta-analyses of the 40-Hz ASSR in patients with ScZ showed trends toward larger effect sizes with stimulus type (click trains > amplitude-modulated tones), analysis level (source > sensor), and stimulus duration (short [≤ 500 ms] > long [≥ 1000 ms]); however, these differences were not statistically significant (Thuné et al., 2016) or substantiated (Zouaoui et al., 2023).The influence of ISI was not examined in the previous meta-analyses.Based on the basic understanding of the role of the 40-Hz ASSR in reflecting cognitive processes and its impairment in ScZ, this meta-analysis aims to investigate the experimental parameters that influence ASSR measurements, with a particular focus on the influence of stimulus duration and ISI.Given the established link between impairments in gamma oscillations and cognitive deficits in brain disorders, including ScZ (Uhlhaas and Singer, 2006), understanding how stimulus duration and ISI affect the 40-Hz ASSR measurements may provide insight into the neural dysfunction characteristic of these disorders.

Methods
This review was conducted following the PRISMA guidelines (Page et al., 2021), and the protocol was registered with PROSPERO (ID No. CRD42024524692).

Search strategy and selection criteria
A literature search was conducted on PubMed and Scopus from November 1999 to December 2023 using the following search terms: ((schizophrenia) OR (psychosis)) AND (steady state response) in PubMed, and TITLE-ABS-KEY ("schizophrenia" OR "psychosis") AND TITLE-ABS-KEY ("steady state response") in Scopus.Inclusion of studies was based on the following criteria: (1) published in a peer-reviewed journal in English; (2) compared patients with ScZ using EEG or MEG to measure 40-Hz ASSRs; (3) reported measures of evoked spectral power and/or intertrial phase coherence (ITPC, also referred to as phaselocking factor); and (4) contained sufficient statistical information (sample sizes and/or mean values and/or standard deviations and/or pvalues and/or effect sizes).When the necessary data for the metaanalysis were unavailable in the publications, we contacted the authors via email to request the data.Exclusion criteria comprised book chapters, reviews, editorials, comments, case reports, conference papers, meta-analyses, qualitative studies, and unpublished studies.We collected available information on age, sex, and stimulus characteristics such as stimulus duration, ISI, stimulus frequency, and sound pressure level.Two raters (SS and KO) independently verified the data validity across all included articles.A flowchart illustrating the selection process is shown in Fig. 1.

Statistical analysis
Effect size calculations were performed using the Comprehensive Meta-Analysis software (version 4).Hedges' g effect sizes were calculated for each measure (spectral power and phase measures) in each study using the sample size along with the mean and standard deviation or p-value or Cohen's d value for each group.As for the studies included in the previous version of the meta-analysis (Thuné et al., 2016), the calculated Hedges' g values were used.The effect sizes were plotted and assessed heterogeneity using Cochran's Q test and І 2 statistics.The effect sizes and 95% confidence intervals (CIs) were estimated under the random-effects model if there was evidence of heterogeneity (p < 0.10).Otherwise, the fixed-effects model was used (p > 0.10).The overall effect size was estimated for spectral power and phase-locking measures for patients with ScZ.Effect sizes are typically categorized as small (g = 0.2), medium (g = 0.5) or large (g = 0.8).To control for publication bias, effect size data were visually inspected using funnel plots (Sterne et al., 2011).The degree of asymmetry in the funnel plots was assessed using Egger regression (Egger et al., 1997).As a post hoc evaluation of the influence of ISI, we defined the ISI as either short (mean, ≤ 1500 ms) or long (mean, ≥ 2000 ms).A previous study investigating the influence of ISI compared three ISI conditions (500, 2000, and 3500 ms) (Choi et al., 2023).Based on this approach, we divided two groups at an ISI of 2000 ms.These categories were treated as two levels of a categorical variable.One-way analysis of variance (ANOVA) was used to compare studies with short ISIs to those with long ISIs, with effect size as the dependent variable.For stimulus duration, similar to the previous meta-analysis (Thuné et al., 2016), we defined stimulus duration as either short (≤500 ms) or long (≥1000 ms) and performed a one-way ANOVA.All statistical analyses were performed at p < 0.05.

Study characteristics
As shown in Fig. 1, the literature searches identified 261 records in PubMed and 164 in Scopus, leading to a total of 425 records.After excluding duplicate records, non-English records, and studies such as reviews and meta-analyses, 226 titles and abstracts were screened.At this stage, 159 reports were excluded, and the remaining 67 full texts were assessed for eligibility based on the criteria.Eight studies did not compare patients with ScZ.Eleven studies did not report either spectral power or phase synchrony.Thirteen studies were duplicates of studies included in the previous meta-analysis (Thuné et al., 2016).Two studies did not provide sufficient data despite our requests.Among studies with overlapping samples, study with the largest number of participants was selected.As a result, in addition to the 17 new studies, 2 studies included in the previous meta-analysis were excluded (Kirihara et al., 2012;Tada et al., 2016).Thus, a total of 33 studies, including 18 studies from the previous meta-analysis and 15 new studies, were included in this metaanalysis.
The combined samples included a total of 1666 patients with ScZ and 1472 HCs.Among the 33 studies, 26 focused on patients with chronic ScZ, and 7 on patients with first-episode psychosis.In addition, 28 studies measured 40-Hz ASSR spectral power, 26 measured 40-Hz ASSR phase-locking, and 21 measured both.
Some studies reported measurements in multiple conditions: three conditions of ISI (500, 2000, and 3500 ms) (Choi et al., 2023), three conditions of sound pressure (75, 80, and 85 dB) (Coffman et al., 2022), two conditions of sound type (click stimuli and amplitude-modulated tones) (Griskova-Bulanova et al., 2018), two conditions of attention S. Sugiyama et al. and ignorance to stimuli (Coffman et al., 2022), and two conditions of eye-opening and closing (Wang et al., 2018).These conditions were matched to the most common conditions in other studies, and data selection was based on this match.Thus, the conditions selected for each study were 500 ms of ISI, 80 dB of sound pressure, click stimuli, ignorance of stimuli, and eye-opening.One MEG study examined spectral power in four regions and phase-locking in one region (Grent-'t-Jong et al., 2021).Therefore, data from the one region of the right Heschl's gyrus common to both measurements were used.In another MEG study, data were recorded using EEG, but the analysis focused on using MEG data (Edgar et al., 2018).

Discussion
In this meta-analysis, the average Hedges' g random effect sizes were − 0.47 and − 0.43 for power and phase, respectively, indicating moderate effect sizes consistent with previous meta-analyses (Thuné et al., 2016;Zouaoui et al., 2023).This suggests that 40-Hz ASSR impairments in patients with ScZ are consistent across diverse patient samples and paradigm designs.The current meta-analysis also showed differences in ASSR measures based on stimulus duration and ISI.Although short stimulus durations (≤500 ms) exhibited larger effect sizes than long stimulus durations (≥1000 ms), consistent with previous meta-analyses, these differences did not reach statistical significance.Conversely, we revealed for the first time that patients with ScZ exhibit significantly larger effect sizes for the 40-Hz ASSR at short ISIs (≤ 1500 ms) compared to long ISIs (≥2000 ms).Furthermore, for spectral power, the effect remained statistically significant after adjusting for stimulus duration, suggesting that the influence of ISI is more pronounced for power than for phase.However, the observation that longer stimulus durations and ISIs reduce the 40-Hz ASSR differences between HCs and patients with ScZ prompts further investigation into how these parameters affect the 40-Hz ASSR.

How stimulus duration and inter-stimulus interval affect the 40-Hz ASSR
Hamm et al. examined the effect of stimulus duration and reported that in HCs, the peak of spectral power approximately 200 ms after the stimulus onset was smaller in the longer stimulus condition compared to the shorter condition (see Fig. 2 in Hamm et al., 2015).On the other hand, Choi et al. investigated the effect of ISI and found similar results for both power and phase (see Fig. 1 in Choi et al., 2023).This suggests that the peaks of the power and phase measures are already determined around the stimulus onset.Thus, the 40-Hz ASSR is likely to be influenced by the history of sensory inputs.
Sensory memory influences subsequent sensory processing, detecting novel sensory memory input (Näätänen et al., 2005;Inui et al., 2010aInui et al., , 2010b)).This system compares previous and present sensory states and triggers novelty-related cortical activity, which is seen as increased auditory evoked potential components such as the N1 (Inui et al., 2010a).The strength of sensory memory, which depends on factors such as the length and decay time of the preceding stimulus and its physical difference from the current stimulus, affects the magnitude of this activity (Inui et al., 2010a(Inui et al., , 2010b;;Ohoyama et al., 2012;Sugiyama et al., 2020).For different stimuli, longer preceding durations and shorter ISIs enhance novelty-related activity (Fig. 3A).For identical stimuli, the opposite is true (Fig. 3B).The magnitude of the N1, which reflects novelty-related activity, is larger for shorter stimulus durations or longer ISIs (Nishihara et al., 2011).We considered that this might explain the variability in ASSR differences between HCs and ScZ patients.Therefore, we examined the relationship between our metaanalysis results and novelty-related activities.
Of the 33 studies in our meta-analysis, 5 used random presentation of different stimuli, whereas 28 used repeated presentation of the same stimuli (Table 1).Of the five random presentation studies, four showed no significant reduction in the 40-Hz ASSR in ScZ patients (Hong et al., 2004;Hamm et al., 2012;Coffman et al., 2022;Nakanishi et al., 2023).Three of these studies used short ISIs (Hong et al., 2004;Coffman et al., 2022;Nakanishi et al., 2023), likely increasing novelty-related activity (Fig. 3Aa).The only random presentation study that showed a significant reduction (Hamm et al., 2015) used a long ISI (~3000 ms), which reduced sensory memory and novelty-related activity (Fig. 3Aa).This suggests that greater novelty-related activity makes it more difficult to discriminate between ScZ patients and HCs.Of the 28 studies using repetitive presentations, only three found no reduction in 40 Hz ASSR in ScZ patients compared to HCs (Edgar et al., 2018;Kim et al., 2019;Bartolomeo et al., 2019).Notably, two of these studies used long ISIs of 2000-6000 ms (Edgar et al., 2018) and 3050-3500 ms (Kim et al., 2019), which increased novelty-related activity (Fig. 3Ba), supporting the idea that large novelty-related activity blurs the difference between HCs and ScZ patients.Different combinations of ISI and stimulus duration complicate the prediction of results.However, the concept that novelty-related activity, influenced by stimulus duration and ISI, affects the magnitude of the 40-Hz ASSR explains most findings where no reduction was observed in ScZ patients.
In a study by Hamm et al. (2015), longer stimuli, which increase novelty-related activity with random presentation (Fig. 3Ab), reduced the 40-Hz ASSR in HCs but not in ScZ patients.Similarly, in a study by Choi et al. (2023), longer ISIs, which increase novelty-related activity with repetitive presentation (Fig. 3Ba), decreased the 40-Hz ASSR in HCs but not in ScZ patients.These results suggest that increased noveltyrelated activity decreases the 40-Hz ASSR in HCs but not ScZ patients.

Hypothesis
Fig. 4 illustrates our hypothesis explaining the effect of stimulus duration and ISI on ASSR and the differences between HCs and ScZ patients.The first key point is that ScZ patients show impairments in novelty detection, which is well documented in studies using mismatch negativity (MMN) (Näätänen and Kähkönen, 2009).Repeated stimuli reduce the responses of cortical sensory neurons that anticipate the next identical stimulus.An unexpected stimulus elicits a larger response, but this response is absent in ScZ patients (Näätänen and Kähkönen, 2009).In addition, auditory-evoked N1 increases with longer ISIs in HCs but not in ScZ patients (Rosburg et al., 2008).For example, significant N1 differences between ScZ patients and HCs were reported for the two longest ISIs (Shelley et al., 1999), suggesting that ScZ patients are less sensitive to ISI changes.The auditory N1 reflects novelty-related activity (Nishihara et al., 2011;ElSayed et al., 2023).
The second key point is that novelty-related activity decreases ASSR in HCs.The suppression of oscillatory activity by novel sensory events, known as phase resetting, was first demonstrated by Rohrbaugh et al. (1989).Abrupt changes during repetitive 40 Hz stimulation modulate the ongoing 40 Hz ASSR for approximately 300 ms (Sugiyama et al., 2021).New sensory events can also accelerate oscillatory activity (Sugiyama et al., 2019a(Sugiyama et al., , 2019b)).Thus, new sensory events affect neural oscillations.The phase deviation of the 40 Hz ASSR correlates with the novelty-related N1 amplitude (Motomura et al., 2019).Furthermore, under conditions with greater novelty-related activity at onset, HCs show a gradual ASSR onset slope, which is even more gradual in ScZ patients (see Fig. 2 in Hamm et al., 2015 andFig. 1 in Choi et al., 2023).Since ASSR is suppressed for 200-300 ms at stimulus onset (Ross et al., 2002) or change detection (Sugiyama et al., 2021), it is suggested that the activity to be suppressed is already smaller in ScZ patients.
This hypothesis is similar to the refractoriness model, which hypothesizes that the recovery time of auditory responses is prolonged due to the refractory period of neural generators (Rosburg and Mager, 2021).Choi et al. proposed this model (Choi et al., 2023), and noted that it is particularly relevant for patients with ScZ, as they are known to have longer refractory periods than HCs (Bickel and Javitt, 2009;Teichert et al., 2016).However, this model suggests that refractoriness increases with more frequent stimulation, which does not explain why the largest ASSR was elicited in HCs under the shortest ISI conditions, as shown in the study by Choi et al. (Choi et al., 2023).
Therefore, our hypothesis is expected to reveal the potential effects of ISI and stimulus duration on ASSR, particularly in distinguishing between HCs and ScZ patients.However, it must be acknowledged that this hypothesis is very preliminary, as the extent to which novelty-Fig.3. Models illustrating the relationship between sensory memory and novelty-related cortical activity.The vertical and horizontal axes represent sensory memory strength and time, respectively.These models are based on previous findings indicating that a single presentation of a sound establishes sensory memory, and the temporal representation of both the storage and decay of sensory memory is logarithmic (Inui et al., 2010a(Inui et al., , 2010b)).Sensory memory is formed during sound presentation and decays during the ISI, and the amplitude of novelty-related cortical activity, reflected by an increase in auditory evoked potential components at stimulus onset such as the N1, is linear to the strength of memory at the onset of the subsequent stimulus.(A) Models in random presentation.Increased sensory memory for the different preceding sounds leads to greater novelty-related activity at the onset of the subsequent sound with shorter ISIs (Aa) and longer stimulus durations (Ab).(B) Models in repetitive presentation.Increased sensory memory for the same preceding sound leads to less novelty-related activity at the onset of the subsequent sound with shorter ISIs (Ba) and longer stimulus durations (Bb).In addition, under random presentations, novelty-related activity increases compared to repetitive presentations, reflecting the physical difference from the preceding sound.ISI, inter-stimulus interval.related activity, dependent on stimulus duration and ISI, influences the ASSR remains unclear.In contrast to the study by Choi et al. (2023), Ross et al. systematically varied the ISI using conditions of 0.5, 1, and 3 s and reported that the amplitude of the ASSR was little affected by the ISI (Ross et al., 2002).It should be noted, however, that the stimulus duration in this study varied across the three ISI conditions.In addition, Ross et al. (2002) proposed that the ASSR can be divided into a "transient gamma-band response" (occurring within 0-100 ms) and a "steady-state response" (occurring after 200-250 ms) based on the distinct stimulus-dependent characteristics of these components.The amplitude of the transient gamma-band response was shown to be influenced by ISI.Therefore, it may be important to observe the early components of the ASSR.More focused studies are needed to understand the effect of novelty-related activity on the ASSR.Moreover, stimulus duration and ISI may influence baseline differences.Kim et al. mentioned that 40-Hz ASSR impairments in patients with ScZ were only observed when the data were not normalized to baseline, suggesting the possibility of increased noise levels (Kim et al., 2019).A reduction in signal-to-noise ratio has been reported in patients with ScZ (Winterer et al., 2000).Spencer reported an increase in total power at 40 Hz in the left auditory cortex in the pre-stimulus baseline period during the 40-Hz ASSR (Spencer, 2012).Furthermore, an increase in the spontaneous power of gamma oscillations, derived by subtracting the evoked power from the total power during the 40-Hz ASSR period, has also been shown in patients with ScZ (Hirano et al., 2015;Tada et al., 2023).Baseline noise or ongoing spontaneous gamma oscillations may not be normalized by the longer stimulus duration and shorter ISI, potentially affecting the 40-Hz ASSR.However, to test this hypothesis, it will be necessary to examine how baseline or background high-frequency activity is affected by stimulus duration and ISI in future studies.

Considerations for future research
To clarify whether the 40-Hz ASSR generation is truly reduced in patients with ScZ, future research must use paradigms with various stimulus durations and ISIs.For example, random presentations with longer stimulus durations and shorter ISIs would increase noveltyrelated activity in HCs, reducing the 40-Hz ASSR (Fig. 4Ab).If these paradigms still fail to show a reduction in the 40-Hz ASSR in patients with ScZ, it cannot be concluded that there is an impairment in the generation of the 40-Hz ASSR.Paradigms that simultaneously assess the 40-Hz ASSR and the N1 would also be beneficial.These paradigms would examine the relation between change detection and ASSR suppression.For example, under repetitive presentations, paradigms with longer stimulus durations and shorter ISIs lead to smaller differences in the N1 between patients with ScZ and HCs, while differences in the 40-Hz ASSR become more pronounced (Fig. 4Bb).Brenner et al. showed in this paradigm that there was no difference in the N1, but a decrease in the 40-Hz ASSR in patients with ScZ (Brenner et al., 2003).However, paradigms with shorter stimulus durations and longer ISIs would result in larger differences in the N1 between patients with ScZ and HCs, with smaller differences in the 40-Hz ASSR (Fig. 4Ba).It is necessary to systematically investigate the relation between change detection and the 40-Hz ASSR using multiple stimulus durations or ISIs.The challenge remains to establish the 40-Hz ASSR as a truly valuable biomarker for ScZ.

Fig. 1 .
Fig. 1.Flowchart detailing the search and selection of reports included in the meta-analysis.
This work was supported by Japan Society for the Promotion of Science KAKENHI (Grant No. 23K14797 [to SS], 23K07026 [to KI], and 22K07614 [to KO]).

Fig. 4 .
Fig. 4. Models explaining the hypothesis that stimulus duration and ISI influence the 40-Hz ASSR.The vertical and horizontal axes represent sensory memory strength and time, respectively.(A) Models in a random presentation that either increase (Aa) or decrease (Ab) the difference in the 40-Hz ASSR between HC and ScZ.(B) Models under repetitive presentation that either decrease (Ba) or increase (Bb) the difference in the 40-Hz ASSR between HC and ScZ.The hypothesis suggests that novelty-related activity suppresses the 40-Hz ASSR in HC, whereas in ScZ, the lack of change in novelty-related activity across different paradigms results in no change in ASSR.Thus, longer stimulus durations and shorter ISIs in random presentation (Ab) and the same parameters in repetitive presentation (Bb) minimize and maximize the difference in 40-Hz ASSR between HC and ScZ, respectively.ASSR, auditory steady-state response; HC, healthy controls; ScZ, patients with schizophrenia; ISI, inter-stimulus interval.

Table 1
Participant and methodological characteristics in each study.