Mismatch negativity in bipolar disorder: A neurophysiological biomarker of intermediate effect?
Introduction
Affecting about 60 million people worldwide, bipolar disorder (BD) is a major mental illness responsible for the loss of more disability-adjusted life-years than all forms of cancer due to its early onset and subsequent chronicity (Merikangas et al., 2011). Characterized by fluctuating mood symptoms, BD is typically expressed by the cycling from major depressive states to hypomanic or manic episodes, which in severe cases are accompanied by psychosis. Early symptoms of BD usually emerge during adolescence and young adulthood (Kim-Cohen et al., 2003, Paus et al., 2008), and if not identified and treated appropriately advance to a chronic state associated with more severe symptoms, greater mood episode frequency and shorter inter-episode intervals (Post, 2007). In this regard BD has been described as a ‘neuroprogressive’ illness (Berk, 2009, Kapczinski et al., 2008, Post, 2007), although the pathophysiology of BD remains poorly understood given its complex and multifactorial nature (Sigitova et al., 2016). In the light of this, there is a need to establish robust biological markers or ‘biomarkers’ of BD to improve diagnostic accuracy and to facilitate preventative strategies (Frey et al., 2013).
Recently, the International Society for Bipolar Disorders (ISBD) Biomarkers Network Task Force proposed some potential candidate biomarkers for BD, with a focus on in three main areas of research: neuroimaging, peripheral biomarkers and genetics (Frey et al., 2013). Specifically, grey matter in cortical-cognitive brain networks, activation in ventral limbic regions and in vivo glutamate levels as indexed by magnetic resonance spectroscopy (MRS) were highlighted as neuroimaging biomarker targets. With regards to peripheral biomarkers: oxidative stress, inflammation and neurotrophins were identified as crucial areas of interest. Finally, genes linked to alterations in calcium metabolism, circadian rhythm, neuronal development as well as brain connectivity were flagged as potential genetic biomarkers of BD. Despite the range and apparent validity of the biomarkers presented, the Task Force noted that there are key unanswered questions relating to the utility of these biomarkers both in terms of predicting outcome (particularly at early stages) and in guiding treatment decisions (Frey et al., 2013).
Thus, it is worth reflecting on the utility of biomarkers in general. As highlighted by Mayeux (2004), while biomarkers often aid in understanding the prediction, cause, diagnosis, progression, regression, or outcome of treatment of disease, their utility “has grown out of the need to have more direct measurement of exposures in the causal pathway of disease” and, while some represent direct steps in the causal pathway of a disease, others are related in some indirect way (Mayeux, 2004). However, Lenzenweger (2013) claims that a biomarker may not be specifically embedded in the causal chain for the disease, but simply reflects “some measurable deviation in the organism, reflective of either internal factors operating in either health/illness or the impact of an external agent”. For Mayeux (2004), there are two major types: biomarkers of exposure (i.e. for risk prediction; indices of the ‘internal dose’ of exposure), and biomarkers of disease (for screening/diagnosis and monitoring progression). Furthermore, biomarkers are often conceptualized as being either trait or state markers, which is particularly pertinent to BD whereby the latter could be useful in the differentiation of mood states or may be applicable only within a specific mood episode (Frey et al., 2013).
Event-related potentials (ERPs), extracted from the electroencephalogram and time-locked to discrete perceptual and/or cognitive events, are commonly utilized as (‘neurophysiological’) biomarkers. In the context of psychiatric disorders, ERPs have been described as particularly important biomarkers (Domjan et al., 2012). One neurophysiological biomarker, mismatch negativity (MMN), has been identified as a ‘breakthrough’ in terms of the understanding and treatment of psychotic disorders (Light and Naatanen, 2013); there are a significant number of studies showing this ERP to be consistently impaired in schizophrenia, with large effect sizes (Erickson et al., 2016, Umbricht and Krljes, 2005). Indeed, with the accumulating evidence (i.e. replication of MMN impairment in schizophrenia samples) there was also an interest in understanding the specificity of this purported biomarker and the natural clinical comparator was often BD. In the sections below, we summarize the literature on MMN in BD (often undertaken in the context of a clinical focus on schizophrenia) and then discuss potential explanations for MMN impairment in BD as reported in recent meta-analyses (albeit to a lesser degree than that observed in schizophrenia) (Chitty et al., 2013, Erickson et al., 2016). First, however, it is important to consider what MMN represents and its underlying mechanism. This is essential as it may be best to conceptualize/utilize MMN within the Research Domain Criteria (RDoC) framework (Insel et al., 2010), whereby it serves as an index of a psychopathology that is shared across psychotic and related disorders, rather than being a diagnosis-specific biomarker (Erickson et al., 2016).
The presentation of a deviant stimulus within a stream of repeated standard stimuli elicits an automatic change detection mechanism, and this can be quantified by measuring the negative going ERP known as MMN (Naatanen, 1990, Naatanen et al., 2007). Accordingly, MMN has also been interpreted as an index of the brain's ability to extract relevant information from an irrelevant background (Hermens et al., 2010). Impairments in deviance detection phenomena, even at the early stages of processing, may induce significant disturbances in higher-order cognitive functioning; hence the importance of such indices in psychiatry and neuroscience (Schmidt et al., 2013). Indeed impairments of MMN have been shown to be significantly associated with functional impairments, in a range of psychiatric samples (Baldeweg and Hirsch, 2015, Hermens et al., 2010, Kaur et al., 2013, Light and Braff, 2005a, Light and Braff, 2005b, Light et al., 2007).
Pharmacologically, MMN is an extremely useful, non-invasive probe of glutamatergic (more specifically, N-methyl-d-aspartate [NMDA] receptor) disturbances due to its reliable attenuation by antagonism at this receptor across animals and humans (Ehrlichman et al., 2008, Heekeren et al., 2008, Javitt et al., 1996, Kreitschmann-Andermahr et al., 2001, Pang and Fowler, 1999, Umbricht et al., 2000). The mechanism explaining the role of the NMDA receptor in MMN has been investigated using dynamic causal modeling, with evidence consistent with the predictive coding hypothesis (Schmidt et al., 2013). That is, the synaptic plasticity necessary for the development of the sensory memory trace is disrupted by NMDA receptor antagonism and therefore MMN is diminished. These findings align with the well-known principal role of the NMDA receptor in regulation of synaptic plasticity within the brain (Bennett, 2000, Bliss and Collingridge, 1993). While there is documented evidence for the roles of dopamine, serotonin (5HT) and gamma-aminobutyric acid (GABA), nicotinic and muscarinic receptors in MMN it is generally accepted that the roles of these agents are less robust and likely exert less regulatory effects (Garrido et al., 2009). A large body of research has shown that antipsychotics tend not modulate MMN (Korostenskaja et al., 2005, Leung et al., 2007, Leung et al., 2010, Pekkonen et al., 2002, Schall et al., 1998, Umbricht et al., 1998, Umbricht et al., 1999) although there is some evidence of an increase in MMN amplitude with antipsychotic treatment (Kahkonen et al., 2001, Zhou et al., 2013). By comparison, the number of studies investigating the effects of other psychotropic medication is scant. In terms of serotonergic modulation, while there is purportedly no change in MMN after 5HT depletion (Leung et al., 2010) or psilocybin (Umbricht et al., 2002), however, the administration of high-dose escitalopram (Wienberg et al., 2010) and tryptophan depletion (Kahkonen et al., 2005) has been shown to increase MMN amplitudes. Studies also have found that MMN does not change with the mood stabilizers, lamotrigine (Vayisoglu et al., 2013) and lithium (Jahshan et al., 2012), nor with anxiolytics and hypnotics (Kasai et al., 2002), benzodiazepines (Murakami et al., 2002), and methylphenidate (Korostenskaja et al., 2008). However, as these findings have not been reproduced, they should be treated with caution.
Disturbances of the NMDA receptor and the neurometabolites involved in its regulation and modulation have been implicated in the pathophysiology of BD (Chitty et al., 2015a, Chitty et al., 2013, Ghasemi et al., 2014, Sanacora et al., 2008). Post-mortem studies showing perturbed NMDA receptor expression, binding, stoichiometry and functioning in BD are commonly detected in the temporal region (Beneyto et al., 2007, Law and Deakin, 2001, McCullumsmith et al., 2007, Nudmamud-Thanoi and Reynolds, 2004, Scarr et al., 2003) and the prefrontal cortex (Beneyto and Meador-Woodruff, 2008, Rao et al., 2010, Rao et al., 2012, Woo et al., 2004). It is noteworthy that many of these findings seem to be specific to NMDA receptors, with no corresponding abnormalities found in other glutamatergic receptors (i.e. AMPA or kainate) (Beneyto et al., 2007, Scarr et al., 2003). Documented success of NMDA receptor antagonists in treating BD also implicates the receptor in the pathophysiology of the disorder. For example, memantine and ketamine, amantadine, D-cycloserine, magnesium and zinc have all shown therapeutic action in BD, and in many cases have shown efficacy in treatment resistant patients (for review see (Ghasemi et al., 2014)).
The earlier studies investigating MMN impairments across clinical groups claimed that MMN impairments are exclusive to patients with schizophrenia; and specifically suggest that it is not impaired in patients with BD (Catts et al., 1995, Hall et al., 2009, Salisbury et al., 2007, Umbricht et al., 2003). On closer inspection of these studies, however, methodological aspects of the respective study designs may warrant caution as the BD samples tended to be smaller in number, rated lower on symptom severity and/or were remitted for longer periods of time compared to the schizophrenia samples (Catts et al., 1995, Hall et al., 2009, Umbricht et al., 2003). In contrast, more recent research has shown that in BD, MMN amplitudes are attenuated and/or have increased latencies (Andersson et al., 2008, Jahshan et al., 2012, Takei et al., 2010). It is interesting to note that in the decade following the null findings reported by the first study of MMN in BD (Catts et al., 1995), there were very few investigations into MMN in BD, despite numerous published studies of MMN in schizophrenia-spectrum subjects. In the following sections, we summarize the BD MMN literature, which has increased substantially over the past decade.
Despite being eight years apart, the first two studies to examine MMN in BD had the same aim: to examine the specificity of this ERP in the context of schizophrenia (Catts et al., 1995, Umbricht et al., 2003). Both studies had small samples of BD (N = 11 and N = 16, respectively) and both found the same result: that MMN (at the frontal midline site, Fz; to duration deviants) in BD was not significantly different from healthy controls (HC), but MMN was significantly reduced in schizophrenia. In the Umbricht et al. (2003) study, it is interesting to note that the proportion of BD individuals (31%) with duration MMN amplitudes 1SD below the normal mean was more than twice as high than in HC (12%) and major depression (14%) groups, and similar to that in the schizophrenia group (35%).
Three subsequent studies also reported no group differences in MMN amplitudes in BD (Hall et al., 2007, Hall et al., 2009, Salisbury et al., 2007). First, Hall et al. (2007) conducted a study utilizing three separate ERP paradigms in twins (16 pairs) and found significant genetic associations between BD and the noted impairments in P300 amplitude as well as P50 suppression, but not MMN; concluding that there is normal genetic architecture related to MMN. These authors replicated their findings in a larger subsequent study of 31 BD families (Hall et al., 2009) concluding that MMN deficits probably occur selectively in chronic schizophrenia. Similarly, Salisbury et al. (2007) found no baseline differences in MMN (Fz) on group comparisons between first-hospitalized schizophrenia, first-hospitalized psychotic BD and HC. However, at longitudinal follow-up (1.7–1.3 years), first-episode schizophrenia (n = 16) but not in first-episode psychotic BD (n = 17) showed a significant reduction in MMN amplitude to frequency deviants (Salisbury et al., 2007). The longitudinal reliability coefficients for the BD group was very low (0.22; cf. 0.74 in schizophrenia) and the study authors noted that this may have been due to the varied patient profiles whereby those who were never re-hospitalized had increased (by 27%) versus those who were re-hospitalized who had decreased (by 20%) MMN amplitudes.
In their study of Bipolar II Disorder (N = 25), Andersson et al. (2008) provided the first evidence that MMN is abnormal in BD, reporting significantly increased MMN latencies and reduced (at Fz) amplitude, compared to HC, in response to duration deviants. Subsequent studies, including some from our group, were found to replicate this finding. Our study (Kaur et al., 2011) of schizophrenia-spectrum and affective-spectrum first episode psychosis provided support for common neuropathological processes affecting deviance detection mechanisms at the onset of both schizophrenia- and affective-spectrum psychoses. Importantly, this study provided evidence of reduced fronto-central MMN in 28 affective-spectrum first-episode patients (n = 10 depression with psychotic features; n = 8 bipolar with psychotic features). In a subsequent MMN study (Kaur et al., 2012a), we specifically compared young patients at the early stages of schizophrenia- (n = 20) versus bipolar-spectrum (n = 20) disorders. Accordingly, patients at early stages of either of these diagnostic spectra showed deficits in fronto-central MMN. In comparison with HC, the schizophrenia-spectrum group showed additional impairments in left temporal MMN. Consistent with our previous studies (Hermens et al., 2010, Kaur et al., 2011), correlational analyses for the entire patient group revealed that reduced MMN was associated with greater psychiatric symptomatology and poorer quality of life. For the first time, the above-mentioned findings suggest that MMN, a purported biomarker of schizophrenia, might index a broader pathophysiology that overlaps traditional diagnostic clusters. In other words, these findings support a shared diathesis model between schizophrenia- and bipolar-spectrum disorders.
Further evidence was provided by Jahshan et al. (2012) who showed their relatively large group of BD (N = 52) to have significantly reduced midline MMN amplitudes compared to HC and intermediate to that observed in schizophrenia (N = 30). Of note, the subsequent P3a component of the MMN ERP was found to be more impaired in BD (and intermediate in schizophrenia). In a study utilising multiple ERP components (including P50, N100, MMN and P300), Domjan et al. (2012) found that their BD sample (N = 20) had significantly increased fronto-central MMN latencies to duration deviants compared to HC. Notably, although the authors reported that the schizophrenia (N = 20) and BD groups showed distinct neurophysiological profiles, visual inspection of the duration deviant MMN waveforms presented suggest that the two patients groups showed responses to duration deviants that were similarly reduced compared to HC (albeit non-significant; p = 0.067).
Following our cross-sectional studies, we examined (Kaur et al., 2013) the predictive utility and longitudinal nature of MMN in early stage schizophrenia- and affective-spectrum disorders. In a sample of N = 28 patients (n = 17 BD), we recorded MMN at baseline and follow-up (19 ± 5 months). Importantly, reduced temporal MMN amplitude at baseline was significantly correlated with greater levels of occupational disability at follow-up, despite diagnosis. Longitudinal assessment revealed that central MMN amplitudes were significantly reduced over time. Those with the smallest MMN amplitudes at baseline tend to have the most severe levels of disability longitudinally, highlighting that MMN may have utility in predicting functional outcomes at early stages of schizophrenia- and affective-spectrum disorders. Furthermore, in such patients, MMN impairments may worsen over time, reflective of ongoing pathophysiological processes.
A more recent study (Baldeweg and Hirsch, 2015) has also examined the specificity of MMN in schizophrenia by including BD and Alzheimer's disease groups as well as HC. The schizophrenia group (N = 49) showed impaired fronto-central MMN that correlated with poor cognitive (in particular memory) performance and this was absent in BD (despite comparable symptomatology and general cognitive deficit), while the Alzheimer's group (who were considerably older) showed impaired memory but intact MMN. According to the study authors, the overall findings provide evidence for a “striking degree of specificity of MMN for the cognitive impairment associated with schizophrenia” (Baldeweg and Hirsch, 2015). Furthermore, the roving MMN paradigm with rapid and constantly changing stimuli exposed a particular weakness of the auditory system in schizophrenia, over and above that observed in MMN's sensitivity to overall cognitive decline (Baldeweg and Hirsch, 2015).
As an alternative to the chronological summary above, Table 1 provides an overview of BD MMN studies ordered according to their effect size. This table lists only studies that included a specific statistical comparison of MMN at frontal and/or central (i.e. including at least Fz or central midline; Cz) sites, between a BD and HC sample). Accordingly, Hedge's g was determined for each study and utilized for comparison, from largest (− 0.98) to smallest (− 0.14) effect (mean = − 0.47; Table 1).
Following our traditional group-based studies, we utilized a data-driven and hypothesis-generating method (i.e. cluster analysis) to determine subgroups, based on MMN profiles, of young patients with schizophrenia- (n = 32) and bipolar- (n = 26) spectrum disorders as well as major depressive disorders (n = 19) (Kaur et al., 2012b). Three clusters were determined: (i) a ‘globally impaired’ cluster (n = 53), characterized by the smallest frontal and temporal MMN amplitudes; (ii) the ‘largest frontal MMN’ cluster (n = 17) distinguished by having the largest frontal MMN amplitudes; and (iii) the ‘largest temporal MMN’ cluster (n = 17) which displayed the largest temporal MMN amplitudes. The three cluster-subgroups were not differentiated in terms of their clinical features (i.e. they showed no differences in psychiatric symptoms). Of note, BD was represented fairly equally across the clusters (25%, 35% and 41%, respectively). However, 55% of the globally impaired cluster consisted of schizophrenia-spectrum patients and this cluster was the most neuropsychologically impaired. The findings of this study suggest that distinct neurophysiological profiles or ‘biotypes’ that cross diagnostic boundaries exist in patients with emerging psychiatric disorders. More research is needed to understand the links between such biotypes and diagnostic categories and/or clinical symptom ratings. In other words, it will be important to determine whether these biotypes simply represent homogeneous underlying pathophysiologies that are separate to traditional diagnostic nosology or whether such findings are due to other factors such as sample sizes or diagnostic uncertainty (given the early age of subjects - around 22 years old); although our finding that the three cluster groups (biotypes) did not differ in various symptom scales suggests that the former explanation may be upheld (and such data-driven analyses will continue to provide new ways to subgroup patients).
Our cluster analysis study (Kaur et al., 2012b) corroborated the results of the only other cluster analytic study of MMN; published the same year (Hall et al., 2012). Despite having a very similar goal to our study, there were some substantive methodological differences in the cluster analysis by Hall et al. (2012). The primary difference is the inclusion of HC (N = 230) and unaffected patient relatives (N = 83). With two patients groups: 68 schizophrenia-spectrum (i.e. schizophrenia and schizoaffective) and 89 psychotic BD, the overall sample size was large, hence the k-means approach (cf. our hierarchical approach). Another key difference was that the Hall et al. cluster analysis utilized multiple ERP components (across three different paradigms) including MMN, P50 sensory gating, N100 and P300. A third difference was the age of the subjects evaluated, with patient groups being at least 15 years older (on average) than the patients groups in our cluster analysis. Notably, the findings were similar to ours in that Hall et al. found distinct neurophysiological profiles that crossed diagnostic boundaries in chronic schizophrenia and psychotic bipolar patients. As revealed in our study, a three-cluster solution was obtained with: (i) a ‘globally impaired’ subgroup; (ii) a ‘sensory processing’ subgroup (with increased N100 or MMN amplitude); and (iii) a ‘high cognitive’ subgroup (characterized by increased P200/P300 amplitudes) (Hall et al., 2012). There were significantly higher proportions of individuals with schizophrenia and BD (but not different from each other) in the globally impaired cluster compared to the other two clusters.
While approaches like cluster analysis provide data-driven, hypothesis generating outcomes, meta-analyses provide crucial information about the consistency of findings across studies. Thus, we undertook a systematic review and meta-analysis (Chitty et al., 2013) of MMN BD studies to determine whether earlier studies that included patients in remission (Catts et al., 1995, Umbricht et al., 2003) or unmatched HC groups (Hall et al., 2007, Hall et al., 2009) may have confounded our understanding of this biomarker. Our meta-analysis (Chitty et al., 2013) included seven studies (Andersson et al., 2008, Catts et al., 1995, Hall et al., 2009, Jahshan et al., 2012, Kaur et al., 2012a, Salisbury et al., 2007, Umbricht et al., 2003) with a total of 204 BD patients (53% female) and 336 HC (60% female) included in the analysis. Overall, there were significantly reduced MMN amplitudes in patients compared to HC (Hedge's g = − 0.42), suggesting that MMN (at least fronto-centrally) is moderately impaired in BD (Chitty et al., 2013).
A very recent MMN meta-analysis (Erickson et al., 2016) focused on schizophrenia patients (101 samples), across a range of stages of illness, in order to examine the progressive nature of MMN impairment. The meta-analysis also included three ‘related conditions’: BD (9 samples1), unaffected first-degree relatives (8 samples) and clinical high risk (CHR; 16 samples), to ‘expand’ on previous meta-analytic findings (in schizophrenia). The effect sizes across groups appeared to support a progression of illness model, with chronic schizophrenia (0.81) having a higher effect size compared to first-episode subjects (0.42) and those at CHR (0.37). However, a larger group labeled ‘schizophrenia-all’ (comprising 75 samples that could not be characterized according to stage of illness) showed the largest effect size (0.91), which led the authors to conclude: “progression in MMN impairment among schizophrenia patients does not follow a linear trajectory”. Pertinent to the current review, the BD samples exhibited an average effect size of 0.37, that is, similar to that of CHR and first episode schizophrenia. The authors offered two potential interpretations of these findings: first, this may reflect a shared neurobiology across disorders; second, that BD sits halfway between HC subjects and schizophrenia patients, and this may be due to the heterogeneity of BD (i.e. those with a history of psychosis may exhibit larger MMN deficits) (Erickson et al., 2016).
Critically, the effect sizes (i.e. Hedge's g ≅ 0.40) for MMN impairment - specific to BD - as determined by two meta-analyses (Chitty et al., 2013, Erickson et al., 2016) are considerably smaller than that (i.e. a mean Cohen's d = 0.99) reported in a schizophrenia MMN meta-analysis of 36 studies (Umbricht and Krljes, 2005). Such differences suggest that the impairments in BD are not as pronounced, or may reflect the substantially lower number of available studies (and subject numbers) in BD.
Two studies utilising magnetoencephalography (MEG) have reported disturbed MMN in BD samples. Firstly, Takei et al. (2010) showed that the mismatch field or ‘MMNm’ (the magnetic counterpart of MMN) response to duration deviants in N = 10 BD patients was significantly delayed compared to HC. Next, Shimano et al. (2014) reported a significantly reduced MMNm in N = 22 BD subjects compared to HC (this time in response to frequency deviants). Notably, there appeared to be an association between the degree of MMNm impairment and illness severity as smaller amplitudes of MMNm were associated with both increased levels of manic symptoms as well as likelihood of previous hospitalisation.
To better understand the neurobiological mechanisms underlying MMN deficits in BD we investigated the relationships between MMN amplitude and the in vivo concentrations of key neurometabolites involved in both the pathophysiology of BD as well as NMDA regulation – glutamate (Chitty et al., 2015a) and glutathione (Chitty et al., 2015b). Specifically, we conducted correlational analyses between frontal and/or temporal MMN and the hippocampal concentrations of these neurometabolites measured via proton magnetic resonance spectroscopy. In both studies, we found a significant association between temporal MMN and each neurometabolite in HC, but not in BD. We interpreted these null findings as suggestive of a perturbed relationship between these measures that may indicate a lack of tightly regulated hippocampal NMDA functioning in BD, or that NMDA receptor regulation in BD is mediated by other factors.
Our group has also conducted studies specifically examining the effects of alcohol misuse on MMN in BD (Chitty et al., 2014, Chitty et al., 2015c). The cross-sectional study involved 42 patients with BD and 34 HC (Chitty et al., 2014) and revealed that both risky drinking and diagnosis of BD predicted impaired temporal MMN. We concluded that this reflected the potential additive effects of alcohol on an already perturbed NMDA/glutamatergic system in BD. Thus, highlighting alcohol as both a modifiable risk factor of neurobiological impairments and as a potential confounder in MMN studies. Twenty-seven of the patients included in the cross-sectional study (Chitty et al., 2014) returned for follow-up, approximately 18 months later (Chitty et al., 2015c). Over this time reduction in risky drinking was associated with increased temporal MMN and decreased fronto-central MMN, as identified via correlational analysis. These longitudinal findings further suggest that risky alcohol use in BD may further compound pre-existing NMDA receptor abnormalities.
Section snippets
Discussion
Despite the claims of the schizophrenia-related specificity of MMN (Baldeweg and Hirsch, 2015, Umbricht et al., 2003), there is mounting evidence in support of the concept that in terms of BD, MMN is a neurophysiological biomarker of intermediate effect. Indeed this is made most apparent by the findings of two meta-analyses revealing moderate effect sizes for MMN impairment in BD (Chitty et al., 2013, Erickson et al., 2016). However, a closer inspection of other studies indicates that there may
Role of the funding source
DFH is supported by grants from the National Health & Medical Research Council (NHMRC) including a Centre of Research Excellence (APP1061043). KMC and MK are both supported by a NHMRC Early Career Fellowship (APP1122362 and APP1112611). The funding source had no further role in the writing of this review and in the decision to submit the paper for publication.
Contributors
All authors contributed to and have approved the manuscript.
Conflict of interest
The authors have no conflicts of interest to declare.
Acknowledgements
We thank Professor Jim Lagopoulos as well as other staff and students at Brain & Mind Centre who have supported the Cognitive Psychophysiology Laboratory.
References (80)
- et al.
Mismatch negativity indexes illness-specific impairments of cortical plasticity in schizophrenia: a comparison with bipolar disorder and Alzheimer's disease
Int. J. Psychophysiol.
(2015) - et al.
Impairment in frontal but not temporal components of mismatch negativity in schizophrenia
Int. J. Psychophysiol.
(2002) The concept of long term potentiation of transmission at synapses
Prog. Neurobiol.
(2000)- et al.
A systematic review and meta-analysis of proton magnetic resonance spectroscopy and mismatch negativity in bipolar disorder
Eur. Neuropsychopharmacol.
(2013) - et al.
Risky alcohol use predicts temporal mismatch negativity impairments in young people with bipolar disorder
Biol. Psychol.
(2014) - et al.
Hippocampal glutamatergic/NMDA receptor functioning in bipolar disorder: a study combining mismatch negativity and proton magnetic resonance spectroscopy
Psychiatry Res.
(2015) - et al.
Investigating the role of glutathione in mismatch negativity: an insight into NMDA receptor disturbances in bipolar disorder
Clin. Neurophysiol.
(2015) - et al.
Different patterns of auditory information processing deficits in chronic schizophrenia and bipolar disorder with psychotic features
Schizophr. Res.
(2012) - et al.
A meta-analysis of mismatch negativity in schizophrenia: from clinical risk to disease specificity and progression
Biol. Psychiatry
(2016) - et al.
The mismatch negativity: a review of underlying mechanisms
Clin. Neurophysiol.
(2009)
The role of NMDA receptors in the pathophysiology and treatment of mood disorders
Neurosci. Biobehav. Rev.
Patterns of deficits in brain function in bipolar disorder and schizophrenia: a cluster analytic study
Psychiatry Res.
Impaired MMN/P3a complex in first-episode psychosis: cognitive and psychosocial associations
Prog. Neuro-Psychopharmacol. Biol. Psychiatry
Effects of haloperidol on selective attention: a combined whole-head MEG and high-resolution EEG study
Neuropsychopharmacology
Serotonergic modulation of mismatch negativity
Psychiatry Res.
Allostatic load in bipolar disorder: implications for pathophysiology and treatment
Neurosci. Biobehav. Rev.
Do high or low doses of anxiolytics and hypnotics affect mismatch negativity in schizophrenic subjects? An EEG and MEG study
Clin. Neurophysiol.
MMN/P3a deficits in first episode psychosis: comparing schizophrenia-spectrum and affective-spectrum subgroups
Schizophr. Res.
Longitudinal associations between mismatch negativity and disability in early schizophrenia- and affective-spectrum disorders
Prog. Neuro-Psychopharmacol. Biol. Psychiatry
Effects of olanzapine on auditory P300 and mismatch negativity (MMN) in schizophrenia spectrum disorders
Prog. Neuro-Psychopharmacol. Biol. Psychiatry
Effect of ketamine on the neuromagnetic mismatch field in healthy humans
Brain Res. Cogn. Brain Res.
Biomarkers: potential uses and limitations
NeuroRx J Am. Soc. Exp. NeuroTherapeutics
Decreased NR1, NR2A, and SAP102 transcript expression in the hippocampus in bipolar disorder
Brain Res.
The effects of benzodiazepines on event-related potential indices of automatic and controlled processing in schizophrenia: a preliminary report
Prog. Neuro-Psychopharmacol. Biol. Psychiatry
The mismatch negativity (MMN) in basic research of central auditory processing: a review
Clin. Neurophysiol.
The NR1 subunit of the glutamate/NMDA receptor in the superior temporal cortex in schizophrenia and affective disorders
Neurosci. Lett.
Kindling and sensitization as models for affective episode recurrence, cyclicity, and tolerance phenomena
Neurosci. Biobehav. Rev.
Dysregulated glutamate and dopamine transporters in postmortem frontal cortex from bipolar and schizophrenic patients
J. Affect. Disord.
Preattentive dysfunction in bipolar disorder: a MEG study using auditory mismatch negativity
Prog. Neuro-Psychopharmacol. Biol. Psychiatry
Mismatch negativity in schizophrenia: a meta-analysis
Schizophr. Res.
Effects of clozapine on auditory event-related potentials in schizophrenia
Biol. Psychiatry
Mismatch negativity predicts psychotic experiences induced by NMDA receptor antagonist in healthy volunteers
Biol. Psychiatry
How specific are deficits in mismatch negativity generation to schizophrenia?
Biol. Psychiatry
Lamotrigine augmentation in patients with schizophrenia who show partial response to clozapine treatment
Schizophr. Res.
Neuropsychological and electrophysiological indices of neurocognitive dysfunction in bipolar II disorder
Bipolar Disord.
Lamina-specific abnormalities of NMDA receptor-associated postsynaptic protein transcripts in the prefrontal cortex in schizophrenia and bipolar disorder
Neuropsychopharmacology
Abnormal glutamate receptor expression in the medial temporal lobe in schizophrenia and mood disorders
Neuropsychopharmacology
Neuroprogression: pathways to progressive brain changes in bipolar disorder
Int. J. Neuropsychopharmacol.
A synaptic model of memory: long-term potentiation in the hippocampus
Nature
Brain potential evidence for an auditory sensory memory deficit in schizophrenia
Am. J. Psychiatry
Cited by (31)
Mismatch negativity and polygenic risk scores for schizophrenia and bipolar disorder
2024, Schizophrenia ResearchElectrophysiological Biomarkers in Dual Pathology
2024, Revista Colombiana de PsiquiatriaGlutamate-based preclinical and clinical dysfunction and treatment in bipolar disorder
2022, Biomarkers in Bipolar DisordersDecreased mismatch negativity and elevated frontal-lateral connectivity in first-episode psychosis
2021, Journal of Psychiatric ResearchCitation Excerpt :MMN impairment may reflect premorbid intellectual functioning in first-episode psychosis as well as a level of vulnerability to disease progression in high-risk psychosis population (Erickson et al., 2016; Salisbury et al., 2017). Furthermore, MMN deficits may not be specific to schizophrenia as they have also been reported in bipolar disorder, autism spectrum disorder, and with mixed evidence for major depressive disorder (Bissonnette et al., 2020; Hermens et al., 2018; Schwartz et al., 2018). This suggests that the neural circuits that produce MMN may be disrupted by a variety of psychiatric disorders.