Intact Wisconsin Card Sorting Test performance: Implications for the role of executive function in schizophrenia

Abstract

The present study sought to examine patients with schizophrenia who are able to perform adequately on the Wisconsin Card Sorting Test (WCST), which is a test that is usually challenging for individuals with schizophrenia due to common problems in executive function. Patients' WCST performance was classified as either 'intact' (fewer than 15% perseverative errors) or 'impaired' (15% or greater). Individuals with intact performance (N=61; 36.3%) and impaired performance (N=107; 63.7%) were then compared on several clinical and socio-demographic variables in order to determine how those with intact executive function differed from those with impaired executive function as measured by the WCST. Patients with intact performance were more similar than different from patients with impaired performance. Only negative symptoms as assessed by Andreasen's Scale for the Assessment of Negative Symptoms (SANS) and Hamilton Depression Rating Scale (HDRS) scores were found to differ significantly between these two groups while the Global Assessment of Functioning Scale (GAF), and positive symptom scores as assessed by Andreasen's Scale for the Assessment of Positive Symptoms (SAPS) approached significance. Patients with schizophrenia with intact WCST performance had a lower mean negative symptom score and a lower mean HDRS score than patients with schizophrenia with impaired WCST performance. The two groups were alike with respect to Beck Depression Inventory (BDI) scores, age at first hospitalization, mean number of previous hospitalizations and their families' years of education. A binary logistic regression indicated that differences between the group of individuals with intact performance and the group of individuals with impaired performance were attributable to full-scale IQ. Those with 'intact' WCST performance exhibited fewer negative symptoms and had lower HDRS scores than those with 'impaired' performance, but both differences were attributable to full-scale IQ based on a binary logistic regression. It is this latter finding that is considered to be the most salient finding of this paper.