Source monitoring improvement in patients with schizophrenia receiving antipsychotic medications.

Rationale: The absence of a relationship between cognitive deficit treatment response and positive symptom treatment response is often assumed, and few data have shed light on this issue. Most of these data have been collected using standard neuropsychological measures, which are ill-designed to assess the types of neurocognitive disturbances associated with psychotic symptoms. This study investigates the effect of treatment on source monitoring performance and its relation to the reduction of certain psychotic symptoms associated with the inability to identify self-generated mental events, known as "autonoetic agnosia".

Objectives: To determine whether risperidone, olanzapine, and haloperidol were differentially effective in reducing autonoetic agnosia and whether changes in this aspect of cognition were related to reduction of specific symptoms of psychosis.

Methods: From a cohort of 49 patients diagnosed with schizophrenia by DSM-IV criteria and randomly assigned to double-blind treatment with risperidone, olanzapine, or haloperidol, 16 patients were identified with symptoms believed to reflect autonoetic agnosia ("target symptoms") as assessed with the Schneiderian Symptom Rating Scale, and then evaluated during a baseline period, and then at 1, 2, and 3 weeks. Autonoetic agnosia was assessed as the ability of a patient to distinguish self-generated words from both experimenter-generated words and pictorially presented words.

Results: Analysis of patients from all treatment groups found a significant reduction in the number of "target" Schneiderian symptoms. Discrimination for items from the self-generated and heard sources significantly improved with treatment, as did the number of self-generated items that patients remembered as coming from the heard source ("self-hear errors"). The correlation between improvement in recognition of self-generated items and reduction in target Schneiderian symptoms after 2 weeks of treatment suggested a modest relationship between symptom improvement and changes in autonoetic agnosia.

Conclusions: While the differences between medications were not statistically significant, antipsychotic medication in general was associated with improvements in symptoms and cognitive deficits that may underlie autonoetic agnosia. Improvement of autonoetic agnosia was a weak predictor of positive symptom improvement in a limited sample.