Cognitive deficit measures in schizophrenia: factor structure and clinical correlates

OBJECTIVE: The authors aimed to delineate the "common denominators" across a representative number of cognitive measures by administering such measures to a large number of schizophrenic subjects. The authors also aimed to study the consistency of the subjects' performance over a 1-year period and to look into the relationship between cognitive task performance and clinical status. METHOD: One hundred fifty-seven schizophrenic subjects with a DSM-III-R diagnosis of schizophrenia were assessed under standardized testing conditions on the dichotic listening and shadowing task, the span of apprehension task, the simple reaction time tasks, the critical stimulus duration, and the interstimulus interval in which they would become free from the masking effect. The clinical status of the subjects was assessed during index assessment, and 97 subjects were reassessed at the end of 1 year. Forty- seven subjects were randomly selected for reassessment on the cognitive tasks administered 1 year earlier. The subjects' performance was compared to that of 56 nonschizophrenic patients with nasopharyngeal carcinoma who had undergone radiation therapy. RESULTS: The cognitive task scores were factor analyzed, giving five distinct factors made up of the span of apprehension accuracy scores, simple reaction time scores, shadowing accuracy scores, span of apprehension error scores, and susceptibility to interference scores. At index assessment, the subjects' shadowing accuracy and reaction time were significantly related to their quality of life, social adjustment, and positive symptoms. The subjects' span of apprehension scores were related to their negative symptoms, measured in terms of work and daily life functions. The initial relationship between the subjects' cognitive test scores and their outcome status remained consistent at follow-up. CONCLUSIONS: The documented cognitive deficits in schizophrenic patients probably reflect a more enduring trait-like status with correlations with outcome status. These results also indicate that the cognitive measures may be simplified in complexity without a corresponding reduction in their clinical significance.