Abstract
Consciousness and hallucinations in schizophrenia: secondary aspects of generalized neuropil pathology?
In his recent essays Bennett reviewed the potential role of synapse pathology in schizophrenia [1, 2]. Following on from Selemon and Goldman-Rakic's neuropil hypothesis of schizophrenia [3], Bennett argued that abnormal synapse formation and regression during childhood and adolescenceresults in de-afferentation in sensory cortices, which, in turn, results in increased spontaneous endogenous neural activity in psychosis [2]. He further argued that these abnormal neural activities give rise to hallucinations due to the lack of corollary discharge; that is, sending an efferent copy from frontal pre-motor areas to sensory cortices, thus allowing an individual to identify such neural activity as self-generated percepts and, in more general terms, gain ‘self awareness’. A failure of this feedforward mechanism inevitably results in a ‘loss of ego boundaries’, a key syndrome of schizophrenia.
Although the notion of impaired corollary discharge in schizophrenia has found some experimental support [4], Bennett's essay falls somewhat short when trying to link auditory hallucinations to the event-related potential measure mismatch negativity (MMN) [2]. MMN is elicited whenever the auditory system detects a change in acoustic stimulation. This process reflects auditory sensory memory processing, which does not require attention, while the detection of a mismatch by the auditory system can trigger attention. Importantly, ‘consciousness’ is not required when recording MMN because the underlying cognitive process is ‘pre-attentive’.
Following the discovery by an Australian research team that MMN amplitudes are reduced in schizophrenia [5], MMN reduction has emerged as one of the most robust findings in schizophrenia research [6]. Reduction of synaptophysin-immunoreactive puncta density in deep layer 3 of area 41 has been proposed to underlie reduced MMN generation [7], thus supporting the notion of synapse pathology in schizophrenia [2, 3] by affecting a specific neural pathway involved in auditory feedforward processing. In contrast, however, MMN amplitude reduction has not been found to be closely associated with auditory hallucinations in schizophrenia, despite a correlation with grey matter decline in primary auditory cortex in the course of illness progression [8].
While positive symptoms – such as auditory hallucinations and persecutory delusions – are common in schizophrenia, they can also occur in a broad spectrum of other mental or neurological conditions or even as a transient phenomenon in the general population without any clinical relevance. Importantly, MMN does not only correlate with grey matter reduction in Heschl's gyrus, it rather extends into the frontal cortex [9], which is consistent with cognitive impairment and decline of global functioning in schizophrenia. Impaired global functioning has also been found to correlate with reduced MMN amplitudes [10]. Hence MMN amplitude measures seem to be more closely associated with Kraepelin's defining aspect of the disorder rather than psychotic phenomena, including auditory hallucinations.
Nevertheless, cortical grey matter decline in temporal lobes is associated with emerging positive symptoms while fronto-parietal grey matter loss appears to correlate with negative symptoms of schizophrenia [11]. As such, the evidence towards a more generalized neuropil pathology in schizophrenia is mounting whereby in vivo structural brain imaging data allow for quantifiable measures of regional brain changes that are associated with the schizophrenia phenotype. Major national research endeavours are currently underway following this strategy, such as the Australian Schizophrenia Research Bank and a large-scale multicentre prodromal study in New South Wales. MMN continues to be an integral part of this scientific journey, thus probing the neural mechanisms underlying schizophrenia.