Social Cognition and Neurocognition in Schizophrenia and Healthy Controls: Intercorrelations of Performance and Effects of Manipulations Aimed at Increasing Task Difficulty

Elizabeth Deckler, Gabrielle E. Hodgins, Amy E. Pinkham, David L. Penn, Philip D. Harvey

Social cognition (SC) and neurocognition appear to predict different aspects of functional outcome in people with schizophrenia. However, the correlations between performance on these domains have not been tested extensively and compared cross-diagnostically with healthy controls. Further, some social cognitive measures appeared to have potential ceiling effects, particularly for healthy people, in previous research, so increasing their difficulty is of interest. In this paper we report on two studies wherein we examined the correlations between neurocognitive ability and performance on SC tests. In the first study the correlations between measures of social perception, emotion processing, and theory of mind and performance on a brief neuropsychological (NP) assessment were examined in 179 schizophrenia (SCZ) patients and 104 healthy controls (HC). In the second study, we instructed participants to perform a subset of the tasks as rapidly as possible in order to increase task difficulty, and we examined the effects of those instructions on task difficulty, task psychometrics, and correlations between SC and NP tests in 218 SCZ patients and 154 HC. In the first study, both HC and SCZ manifested a domain specific pattern of correlation between NP and SC test performance. Controlling for group differences in NP performance did not eliminate SC performance differences between the groups. In the second study, no differences in task performance, intercorrelations other SC tests, or test-retest stability were induced by the difficulty manipulation in the samples who performed the tasks with speed demands compared to the performance of the previous sample. These data suggest that simple manipulations aimed at increasing task difficulty may not have the desired effect and that despite consistent correlations between SC and NP test performance, impairments in social cognitive functioning are not fully explained by NP performance deficits.