ABSTRACT
De®cits in various domains of cognitive functioning have increasingly been recognized over the last decade as a core feature of schizophrenia (Heinrichs, 2004). These not only include de®cient performance in higher cognitive domains but also extend to information processing at the sensory and pre-attentive level. Such cognitive abnormalities can be studied through neurophysiological techniques. Event-related potentials (ERPs) are changes in the human electroencephalogram (EEG) that occur in response to stimuli. They allow measurement of cortical physiology while the subject is performing cognitive tasks and thus provide a non-invasive technique of evaluating neural activation during cognitive processes. Importantly, ERPs are inexpensive and easily obtained from large numbers of subjects, and are stable quantitative measures. The main limitation of neurophysiological techniques to examine cortical activity is their low spatial resolution and thus the precise anatomical sources of the signals are not well identi®ed. However, this is compensated for by an extremely high temporal resolution that enables measurement of changes in the human EEG during cognition and perception in vivo, in real time. Neurophysiological experiments used in psychosis research range from basic pre-pulse inhibition tests (Kumari et al., 2005) to high-order N400 language tasks (Matsumoto et al., 2005). However, it is the P300, mismatch negativity (MMN) and P50 waveforms that have been best characterized as schizophrenia endophenotypes. In this
chapter, we describe each of these waves and present a meta-analysis of existing data on P300 and P50 wave abnormalities in patients with schizophrenia and their unaffected relatives. We then describe data from the Maudsley Family Study of Psychosis, which explored abnormalities of the P300 and MMN waves in patients with schizophrenia and in their unaffected ®rst-degree relatives.
