This study investigated feature binding in a working memory task in patients with schizophrenia and in normal controls. Twenty-five patients and 25 controls participated. On each trial, three drawings of familiar objects were presented sequentially, each in a different cell of a 3=3 grid. In different blocks of trials, participants remembered either individual features (object and location conditions) or an object and its location (combination condition). The results showed that patients were slower and less accurate than controls under all conditions. Accuracy of both groups was reduced in the combination condition relative to the single-feature conditions, but patients showed disproportionately poorer performance in the combination condition than in the object and location conditions. Thus, patients with schizophrenia exhibit deficits in working memory, particularly when the task requires binding objects to their locations. This finding demonstrates that processes that establish coherent and temporary episodic representations in working memory are impaired in schizophrenia. ᮊ

No abstract

Patients with schizophrenia exhibit normal memory for separate objects or locations but are disproportionately impaired when the items must be bound for later recognition in a working memory (WM) setting (Burglen et al., 2004). This study aimed at further evaluating the contribution of each WM component to the patients' binding deficit, using selective articulatory, visuospatial, and executive suppression tasks. In the object-location binding task used, a trial comprised the successive presentation of three drawings of familiar objects and of three spatial locations in a grid, either separately (i.e., objects alone or locations alone) or bound (i.e., object1location), and required a recognition test following an 8-s delay. In the suppression modalities, suppression was continuous from presentation to test. A total of 22 patients with schizophrenia and 24 healthy controls participated. The results confirmed the binding deficit in patients' performance in the baseline modality where no suppression was required. They also showed that patients were particularly disrupted when suppression was visuospatial. This last finding extends the specific visuospatial vulnerability in schizophrenia to the operations of binding.