Semantic memory deficit in schizophrenia

This study is designed to study the extent of semantic memory deficit in schizophrenia, its nature, and its correlates, in community-based schizophrenic patients using a semantic memory battery. The patients’ performance was compared with that of matched controls. About six months later, the same cohort was tested again using the same battery. Patients’ performance was analysed using the criteria proposed by Warrington and Shallice (1979), to decide whether this deficit conforms into an access or store type disorder. The criteria used where: consistency across time, level of attributes and the effect of cueing. Typical parametric analysis showed significant impairment, on average, for this group of patients compared to controls on most tests of this battery. About a quarter or less of patients would perform below that of 5th percentile of the controls group on different tests of the battery used. However, a more robust and atypical analysis strategy demonstrated that the significant impairment is limited to two tests (category fluency and naming to description). In addition, this impairment correlate with negative symptoms scores (on the SANS) but no significant correlation with positive symptoms scores (on the SAPS). Across time, improvement on SANS score correlated with improvement in some semantic memory tests scores (category fluency and naming to description). Further analysis showed patients performance followed the same pattern of controls. This would be interpreted as difficulty in access rather than degraded store on the criteria mentioned above. The strongest of evidence came from patients’ performance on a cueing paradigm and inconsistency. These results are in support of earlier suggestion of an access semantic memory deficit in schizophrenia. The results could be related, in this cohort of patients, to being relatively young, with no significant symptomology, and living in the community.