The antisaccade task and neuropsychological tests of prefrontal cortical integrity in schizophrenia: empirical findings and interpretative considerations.

To date, every published study of the antisaccade task has replicated the finding that schizophrenia patients make an increased number of errors. This finding has been interpreted as support for frontal and/or basal ganglia dysfunction in schizophrenia, primarily because neurological patients with pathology in these brain regions also make large numbers of errors on the antisaccade task. Here, we compared the performance of schizophrenia patients and nonpsychiatric controls on an antisaccade task and on two neuropsychological tests, the Wisconsin Card Sorting Test, which is assumed to tap frontal lobe functioning, and the interference condition of the Stroop Test, which is thought to tap dorsolateral prefrontal cortex/anterior cingulate functioning. We examined the pattern of intercorrelations among these tasks. Schizophrenia patients made significantly more errors on the antisaccade task, made more perseverative errors and achieved fewer categories on the Wisconsin Card Sorting Test, and were significantly slower during the interference condition of the Stroop Test than were nonpsychiatric controls. Antisaccade errors were significantly correlated with interference performance on the Stroop in schizophrenia patients and in controls, but were not significantly correlated with the measures of Wisconsin Card Sorting Test performance in either group. The pattern of intercorrelation suggests that these tasks should not be thought of as representing a unitary variable of "frontal cortical integrity". Although aspects of these tasks tap the ability to inhibit prepotent responses, each task is also behaviorally complex. The multifaceted nature of these tasks makes it difficult to isolate which brain regions are part of the network underlying the specific act of inhibiting a prepotent response (for example, the reflexive saccade toward the novel peripheral target) and which regions participate in aspects of task performance that are related to non-inhibitory components (for example, executing an antisaccade). A broadly distributed network is undoubtedly involved in both processes. Parsing the various components of cognitively complex tasks may help to clarify both the specific behaviors that are anomalous and their underlying neural substrates. We also address the complexity of inferring localized brain dysfunction in schizophrenia patients based on seemingly analogous behavioral deficits in neurological populations.