An incongruent reality: the N400 in relation to psychosis and recovery

Abstract

Background: Cognitive impairments and delusions are hallmarks of schizophrenia, and are thought to be due in part to abnormalities in semantic priming. The N400, a neural measure of semantic processing, is found to be reduced in schizophrenia. However, it is unclear if individuals with other psychoses (e.g., mood disorders or substance abuse with psychotic features) also show this impairment, and whether N400 reduction relates to real-world functioning and recovery.

Methods: Eighty-nine individuals from the Suffolk County Mental Health Project, a longitudinal study of first-admission psychosis, and 35 healthy adults were assessed using matched, related, and unrelated picture-word pairs to elicit the N400. Patients' real-world functioning, symptomatology, and recovery were tracked since first hospitalization; EEG assessment was completed during year 15 of the study.

Results: Participants with schizophrenia had slower reaction times and reduced N400 to semantically incongruent stimuli relative to healthy participants. Schizophrenia and other psychoses did not differ on N400, suggesting that N400 abnormalities characterize psychosis broadly. When grouped by recovery status, patients who remained ill had a significantly blunted N400, while those who recovered did not differ from healthy adults. Few patients with schizophrenia achieved recovery; therefore recovery results are limited to the other psychosis group. Furthermore, reduced N400 and increased reaction times correlated with greater psychotic symptoms, worse global assessment of functioning scores, unemployment, and impaired social functioning.

Conclusions: Abnormalities in the N400 are not specific to schizophrenia; in addition, the N400 may be a useful neural correlate of recovery and real-world functioning across psychotic disorders.

Keywords: ERP; N400; Psychosis; Recovery; Schizophrenia.

Figure 1

The N400 for Healthy Comparison, Other Psychosis, and Schizophrenia groups. Waveforms show an average of electrode sites Cz, Pz, CP1 and CP2. The 200ms immediately before word onset contained brain activity associated with picture processing and therefore was not shown or included in analyses. The pre-stimulus baseline shown (−200 to 0ms) represents the pre-picture baseline used in analysis, which was measured from −525 to −325ms pre-word onset. Headmaps show the topography of the related and unrelated difference waves (related minus match, and unrelated minus match, respectively).

Figure 2

The N400 waveforms for patients who were recovered compared to those who were not recovered. Waveforms show an average of electrode sites Cz, Pz, CP1 and CP2. Headmaps show the topography of the related and unrelated difference waves (related minus match, and unrelated minus match, respectively).