The dynamic architecture of working memory in schizophrenia

Background: The capacity to hold information in working memory is greater for the first and/or last items of a sequence of information (architecture), and varies according to the retention interval (dynamic) and the type of stimuli. Although working memory deficits in schizophrenia have been documented widely, it is not clear how its architecture and dynamics are affected by the disease. Methods: Using two Sternberg paradigms - the recognition and the context-recall tasks - we investigated the effect of serial position, retention interval, type of stimuli, and task (type of encoding for the serial position) on working memory capacity in 26 schizophrenia patients and 20 healthy control subjects. A mixed model analysis of variance was applied to the proportion of correct responses and reaction time data. Results: All the experimental factors had significant effects. However, the most important effects were those of group, group × serial position, and group × delay interactions. The last two effects were driven by a reduced primacy effect and by a reduced performance with longer delay in schizophrenia compared to control subjects. The serial position × delay interaction was significant without triple interaction with group. Group × type of stimuli and group × task for the serial position interactions were not significant. Conclusion: Schizophrenia patients exhibited normal dynamics but abnormal architecture of working memory (reduced primacy effect), and faster decay of information. These impairments affected equally verbal, spatial and object stimuli and operated with implicit and explicit encoding of the serial position. Although these impairments were not correlated with the clinical picture, they are likely to contribute to the pathogenesis of the difficulties with which schizophrenia patients are faced. Consequently, addressing these specific impairments could alleviate these difficulties.