Impaired mismatch negativity generation in prodromal subjects and patients with schizophrenia

Abstract

Background

Mismatch negativity (MMN) specifically the response to tone duration deviants has consistently been shown to be reduced in schizophrenia suggesting dysfunction in auditory sensory memory.

As part of a multidimensional approach to the early recognition of psychosis, MMN was investigated as a possible risk factor for later development of psychosis in subjects with a prodromal syndrome. Forty-three prodromal subjects, 31 neuroleptic-free inpatients with schizophrenia and 33 healthy controls were studied. A prodromal state was defined by a cluster ‘Cognitive Disturbances’ as defined by the ‘Bonn Scale for the Assessment of Basic Symptomsʼ (BSABS), which was found highly predictive of first-episode schizophrenia. To elicit MMN, a three-tone auditory oddball paradigm with 10% ‘duration deviants’ and 10% ‘frequency deviants’ was used.

Results

MMN amplitudes to tone duration deviants were significantly reduced in the patients with schizophrenia compared to controls. The putatively prodromal subjects also showed a slight, though non-significant reduction of the MMN amplitude that was intermediate between normal controls and patients with schizophrenia, and with a larger within-group variance.

Conclusion

These results support the view that abnormalities in temporal processing are particularly pronounced in patients with schizophrenia. Prodromal subjects are a heterogeneous group with regard to outcome and time until transition to a first psychotic episode. Follow-up of these putatively prodromal subjects will show whether MMN amplitudes further reduce over time in those developing psychosis and if a reduction is state-dependent.

Introduction

Deficits in various domains of cognitive functioning are considered a core feature of schizophrenic psychoses (Goldberg and Gold, 1995). Furthermore, recent findings indicate that certain impairments in cognitive functioning precede the onset of the first psychotic episode and are already evident in persons who are considered to be at risk of psychosis (Hambrecht et al., 2002, Wood et al., 2003). To understand the biological basis of these cognitive deficits, the underlying neural correlates need to be investigated. Thus, event-related potentials (ERP) appear to be effective paradigms as they allow a real-time observation of neurophysiological processes while the subject is performing the respective cognitive task. Of the different ERP paradigms, mismatch negativity (MMN), a component of auditory ERP, has consistently been shown to clearly discriminate between patients with schizophrenia and controls (for review, see Mitchie, 2001), thus providing a promising starting-point for the investigation of neurobiological at-risk indicators in potentially prodromal subjects.

MMN has been interpreted as an index of automatic feature analysis in the auditory cortex, the so-called ‘echoic memory’ (Näätänen, 1990), with respect to temporal processing (Davalos et al., 2003). It is elicited as a response to rare deviant tones interspersed in the repeated presentation of a standard tone and is little affected by attention and motivation. While a reduced MMN amplitude in response to tone duration and tone frequency deviants, respectively, has been demonstrated for patients with schizophrenia in several studies (Shelley et al., 1991, Shelley et al., 1999, Javitt et al., 1993, Javitt et al., 1995, Javitt et al., 1998, Catts et al., 1995, Reite et al., 1996, Oades et al., 1996, Shutara et al., 1996), MMN response to tone duration deviants in particular seems to be impaired in schizophrenia (Mitchie et al., 2000, Mitchie, 2001, Umbricht et al., 2003).

To investigate the applicability of MMN as a neurobiological at-risk indicator reflecting an echoic memory disturbance, potentially prodromal subjects (PP), neuroleptic-free patients with schizophrenia (SP) and healthy controls (HC) were compared for the first time. It was assumed that, in comparison to healthy controls, those with schizophrenia would exhibit a large, significant reduction of the MMN amplitude, whereas the MMN amplitude of prodromal subjects would show a less pronounced though still detectable reduction, thereby holding an intermediate position between those with schizophrenia and healthy controls. Moreover, a larger group effect was expected on the MMN amplitude in response to tone duration deviants compared to that in response to tone frequency deviants. In addition, associations between MMN reduction and neuropsychological deficits were investigated to detect possible interactions of these two different levels of neurobiological functioning.