In the SG experiment, we found that patients with schizophrenia showed significantly higher theta power in all six brain regions under the induction of S1 and S2 than healthy controls. The theta power induced by S2 was lower in the top and central regions than in S1; however, there was no significant difference in the power values evoked by patients S1 and S2. In the MMN experiment, patients with schizophrenia exhibited higher theta power in the parietal and temporal regions under low-frequency stimulation, while patients induced higher theta power in the central brain and left temporal regions than healthy controls under high-frequency stimulation. In the SG paradigm, the gamma power induced in patients with schizophrenia was higher than that in the healthy control group. However, in the occipital region, patients with schizophrenia would show lower delta power than healthy controls. In the SG paradigm, the gamma power induced in patients with schizophrenia was higher than that in the healthy control group. However, in the occipital region, patients with schizophrenia would show lower delta power than healthy controls. In the MMN experiment, patients showed higher beta power in the parietal region under both low- and high-frequency stimulation than the control group. In the gamma band, low-frequency stimulation made the patients in the central region higher than the control group. With high frequency stimulation, patients were higher in the parietal region and left temporal.
4.1. Abnormal theta in schizophrenia in both SG and MMN
It was found that patients with schizophrenia showed significantly higher theta power in the SG paradigm than healthy controls, similar to the conclusions mentioned in the literature [5] In many patients with schizophrenia, various evoked potential patterns in vitro and in vivo suggested that stimuli evoked early theta frequency oscillations. Previous studies have reported dysfunction in the left temporal region involved in speech classification in schizophrenia and reduced gray matter volume in the occipital, parietal, frontal, and temporal cortex in schizophrenia compared with healthy controls [7, 15]. Therefore, we analyzed each of the six brain regions, and the results showed that schizophrenic patients showed higher theta oscillations in the whole brain. At the same time, we found that the theta power induced by S2 in healthy controls was lower in the parietal region and central region than that induced by S1, which was consistent with the adaptive phenomenon of the SG paradigm, and there was no difference between the two groups in schizophrenia, indicating that patients with schizophrenia were not well adapted to stimuli, which was reflected in the theta band on the spectrum and only in the central and brain regions. It should be noted that the current studies on SG paradigm tasks in patients with schizophrenia had few studies on the power spectrum to analyze the differences between brain regions. Studies on ERP have found that [16] patients with schizophrenia and healthy controls had significant differences in peaks in the occipito-temporal region, which suggested that the exploration of power differences between different brain regions was a relatively new field, and further research in this field may lead to further discussion of diagnostic indicators for patients with schizophrenia.
In general, studies had shown that MMN neurogenesis was indicative of an ERP component of auditory discrimination [17]. For example, during the frequency recognition of simple tones, the size of the MMN increased as the frequency difference increased between the standard and the deviation [18]. Mismatch negativity was the negative component of the waveform, obtained by subtracting the correlated response to frequent stimuli (standard) and the correlated response (ERP) to rare stimuli (deviation). The MMN could be elicited whether or not the subject noticed the sequence, and it reflected the function of auditory sensory memory. Mismatch negative production deficit have been identified as one of the best potential biomarkers of cognitive impairment in patients with schizophrenia [19]. Many studies have reported the frequency and duration defects of mismatch negatively [20], so we also used the MMN paradigm to analyze the power difference between the two groups. In one study, through the frequency analysis of the oddball paradigm experiment of healthy subjects, it was found that compared with the standard stimuli, the event-related slow potential with biased stimuli seemed to show stronger synchronization in the frequency range of the theta band [21]. Under high-frequency stimulation, that was, biased stimulation, schizophrenic patients had higher theta power than healthy controls, which was consistent with the results of a previous study [2]. Theta oscillations could be induced by biased stimuli, which was consistent with the extensive cognition that theta oscillations reflected auditory cognitive function. However, patients with schizophrenia showed higher theta oscillations than healthy people under the same high-frequency sound stimulation, namely, biased stimulation. This suggested that theta oscillation could be used as one of the potential biological indicators of auditory tests in patients with schizophrenia.
One study found that healthy subjects also induced higher theta oscillations to biased stimuli [21]. In our study, healthy subjects did not show this difference. The author speculates that this may be because of the need to respond to high-frequency sound stimuli in this experiment, so before the beginning of the experiment, the researchers would tell the subjects that high-frequency sound stimuli would occur randomly in this experiment, so the subjects were psychologically prepared for this. This preparation may reduce the brain's response to biased stimuli. Then, we considered whether sound stimulation also induced theta oscillations in the whole brain. Some EEG bands have been shown to be related to different perceptual and neurocognitive processes in speech processing, including theta (4–7 Hz), beta (12–30 Hz) and gamma (30–50 Hz) [17, 22]. Auditory hallucination was one of the typical symptoms of schizophrenia; auditory hallucination originated from the auditory cortex and serves the area of receptive language processing, while auditory information processing disorders occurred repeatedly in schizophrenia. Some studies have observed an increase in theta rhythm in the left superior temporal region of schizophrenic patients during auditory hallucinations [23], which was similar to our research results. Current studies provided consistent evidence that careful examination of theta band activity during speech processing may selectively distinguish between schizophrenia and healthy subjects or other psychiatric disorders.
Therefore, theta oscillation could be used as a focus in the study of schizophrenia, and sound stimulation could always induce an abnormal increase in theta oscillation. When the frequency of sound stimulation was different, the increase in induced theta oscillation was reflected in several specific brain regions, including the temporal lobe. This meant that the abnormal increase in theta oscillation was associated with abnormal auditory processing and auditory hallucination symptoms in patients with schizophrenia.
4.2. The anomaly of theta was more obvious than other frequency bands
One of the well-known components in ERP studies was that P300 reflects people's advanced cognitive functions, such as perception and attention, while delta (0–3 Hz) and theta (4–7 Hz) were the main contributors to P300 [22, 24]. Some studies have suggested that the decrease in P300 in schizophrenia may be related to the decrease in theta activity [25], indicating that P300 was related to theta oscillation. The temporal region was mainly responsible for auditory perception and participates in long-term memory and emotion. Some studies have found that the P300 component amplitude of the left temporal scalp region could be used to distinguish between normal and schizophrenia [26], and the comparison of auditory P300 components of event-related potentials between the normal control group and the schizophrenic group confirmed the left temporal lobe defect of P300 amplitude in early schizophrenia [16]. It was mentioned earlier that theta was one of the main contributions of P300 components, which indirectly proved the finding in our study that patients with schizophrenia induced higher theta power in the temporal region under high frequency stimulation.
Similar to P300 induced by deviant stimulation, the presentation of biased stimulation also induced gamma and theta oscillations in the medial frontal cortex and parietal cortex [27]. This was consistent with the conclusion that there was a significant difference in gamma power between the parietal region and the left temporal region between the healthy control group and the schizophrenic group under high-frequency stimulation. Some studies have excavated deep into the relationship between frequency and P300 and show that the topography, time process and sensitivity to task requirements of delta and theta were consistent with those of P3b and P3a, respectively [28]. At present, there were two sources of delta oscillation, one of which originated from the cortex, which may be related to cognitive function in the awake state. In this study, the schizophrenic group showed higher delta power than the healthy control group under stimulation of S1 and S2. Next, the authors found that few studies focused on whether there were differences in delta oscillations between schizophrenic patients and normal subjects in different brain regions. Therefore, we analyzed the six brain regions one by one and found that there was a significant difference between the patients with schizophrenia in the parietal region and the healthy controls. From this, we inferred that patients with schizophrenia have certain cognitive defects in the parietal region of the brain.
There were two main theories about the mechanism of MMN. One of the theories was ‘neural adaptation’, which explained that repetitive standard stimuli leaded to adaptation and attenuation of neural activity [8]. The neural adaptation was consistent with the adaptation phenomenon in sensory gating. Sensory gating was a normal function of the brain that could specifically inhibit irrelevant and redundant sensory stimuli inputs. The inhibition of SG P50 was considered to be a reflection of the brain's inhibition mechanism of information overload, which was usually evaluated by the auditory paired click paradigm. Many studies have shown an increase in SG gating in patients with schizophrenia [29, 30]. At the same time, in the time-frequency domain, the main contribution of the auditory P50 ERP response was in the frequency range of gamma and theta. In one study, the gamma band power of schizophrenic patients and healthy controls during the oddball paradigm was evaluated [31], and it was found that the power of P3, P4, Fz, Pz and T5 in the patient group was higher than that in the healthy control group, and P3, P4, Pz and T5 were located in the parietal region and temporal region, respectively. Our study extended from the electrode site to the brain region and found that schizophrenic patients in the parietal, left temporal and central regions showed higher gamma power than healthy controls. The interaction between two oscillations in different frequency bands was called coupling. In phase-amplitude coupling, the amplitude of the high-frequency oscillation was modulated by the phase of the low-frequency oscillation. For example, theta-gamma coupling, where the phase of the theta oscillation adjusted the amplitude of the gamma oscillation. In our results, gamma and theta showed similar results; that was, compared with healthy controls, schizophrenic patients induced higher levels of oscillations in the parietal, central, and temporal regions.
In summary, the abnormal oscillations of theta, gamma and delta in different brain regions induced by sound stimulation may be used as an early index for the diagnosis of schizophrenia. Compared with other oscillations, the abnormality of theta oscillation was specific and stable. It also showed that theta oscillation could be used as a focus and biomarker in the study of schizophrenia.