Abnormal ERPs and Brain Dynamics Mediate Basic Self Disturbance in Schizophrenia: A Review of EEG and MEG Studies

Abstract

Background: Interest in disordered sense of self in schizophrenia has recently re-emerged in the literature. It has been proposed that there is a basic self disturbance, underlying the diagnostic symptoms of schizophrenia, in which the person's sense of being a bounded individual continuous through time loses stability. This disturbance has been documented phenomenologically and at the level of cognitive tasks. However, the neural correlates of basic self disorder in schizophrenia are poorly understood. Methods: A search of PubMed was used to identify studies on self and schizophrenia that reported EEG or MEG data. Results: Thirty-three studies were identified, 32 using EEG and one using MEG. Their operationalizations of the self were divided into six paradigms: self-monitoring for errors, proprioception, self-other integration, self-referential processing, aberrant salience, and source monitoring. Participants with schizophrenia were less accurate on self-referential processing tasks and had slower response times across most studies. Event-related potential amplitudes differed across many early and late components, with reduced N100 suppression in source monitoring paradigms being the most replicated finding. Several studies found differences in one or more frequency band, but no coherent overall finding emerged in this area. Various other measures of brain dynamics also showed differences in single studies. Only some of the study designs were adequate to establish a causal relationship between the self and EEG or MEG measures. Conclusion: The broad range of changes suggests a global self disturbance at the neuronal level, possibly carried over from the resting state. Further studies that successfully isolate self-related effects are warranted to better understand the temporal-dynamic and spatial-topographic basis of self disorder and its relationship to basic self disturbance on the phenomenological level.

Keywords: basic self; electroencephalography; event-related potentials; magnetoencepalography; schizophrenia; self disturbance; sense of self.

Figure 1

Studies divided by experimental paradigm. The studies that were selected for inclusion in the literature review grouped into six paradigms based on the type of self-related task used in their experimental design.

Figure 2

Changes in behavioral responses. The studies grouped into the same paradigms as in Figure 1, showing the differences between subjects with schizophrenia and healthy controls in task responses. (A) Differences in accuracy. (B) Differences in response times.

Figure 3

Changes in ERP component amplitudes including null results. Studies that compared ERP component amplitudes in subjects with schizophrenia to those in healthy controls. SCZ, subjects with schizophrenia. HCs, healthy controls. Multipliers beside a plus, equals, or minus sign indicate the number of studies that reported this result. Studies marked as “self/other” compared self-related task data to comparable non-self-related task data to establish a role for the self. Studies marked as “self only” used a self-related task but without a control condition to establish that it was this aspect of the task that caused the ERP component change. The visual depictions of ERP components are stylized approximations that do not correspond to the findings of any of the studies in particular. A few ERP components were not named in the studies and were categorized here based on their latency and whether they were positive- or negative-going. Results of the individual studies can be found in the Supplementary Material.

Figure 4

Changes in ERP component amplitudes by paradigm. Studies that compared ERP component amplitudes in schizophrenia to those in healthy controls, excluding those that did not compare self-related task data to comparable non-self-related task data. Sorted according to the paradigms from Figure 1. SCZ, subjects with schizophrenia. HCs, healthy controls. Multipliers beside a plus, equals, or minus sign indicate the number of studies that reported this result. The visual depictions of ERP components are stylized approximations that do not correspond to the findings of any of the studies in particular. A few ERP components were not named in the studies and were categorized here based on their latency and whether they were positive- or negative-going. Results of the individual studies can be found in the Supplementary Material.

Figure 5

Changes in individual frequency bands. Results comparing individual frequency bands in subjects with schizophrenia to those in healthy controls. SCZ, subjects with schizophrenia. HCs, healthy controls. FTC, frontotemporal coherence. ITC, intertrial coherence. PLI, phase lag index. PLF, phase-locking factor. The presence of multiple signs beside one measure indicates multiple separate results. Studies marked as “self/other” either compared self-related task data to similar non-self-related task data or else correlated their findings with a psychopathology scale that measures self disorder. By contrast, studies marked as “self only” used neither of those methods to establish whether the observed change was related to the self. Most of the studies included in this figure used a source monitoring paradigm. However, the following results were from other paradigms. Self-referential processing paradigms were used by Jia et al. (25), who found no difference in theta or beta power but decreased alpha power and phase lag index. Proprioception paradigms were used by Arnfred et al. (22), who found decreased beta amplitude and phase-locking factor, increased gamma amplitude, and no difference in gamma phase-locking factor, and Arnfred et al. (23), who found a positive correlation between beta amplitude and psychopathology scales but a negative correlation between gamma frequency and psychopathology scales.

Figure 6

Changes in other dynamic measures. Dynamic measures studied that were not individual frequency bands, labeled with the paradigm used to obtain them. SCZ, subjects with schizophrenia. HCs, healthy controls. With the exception of the study by Kim et al. (16), all of these studies differentiated self-related effects from non-self-related effects either through comparison of experimental conditions or through correlational analyses with psychopathology scales that measure self disorder.

Figure 7

Changes in resting state and prestimulus measures. Findings on the resting state and prestimulus period, including comparisons of subjects with schizophrenia to healthy controls and correlations with psychopathology scales. SCZ, subjects with schizophrenia. HCs, healthy controls. PLF, phase-locking factor. Neural integration (47) was computed as correlations between 16 electrode pairs. For the resting state findings, a direct link to the self was established only by Bob et al. (47), by correlation with the Dissociative Experiences Scale. For the prestimulus findings, a direct link to the self was established only by Ford et al. (2007) in their comparison of patients and controls, through an interaction effect with experimental condition. This figure does not show findings from studies examining rest/task differences or prestimulus/task differences, as these were shown in Figure 6.

Figure 8

Proposed relationships between empirical findings. The major groups of findings identified in this review. Thick, solid arrows indicate relationships that were studied empirically in the literature while thin, dashed arrows indicate relationships that the basic model of self proposes in order to unify separate findings in the literature.