RÉSUMÉ
OBJECTIVES: This study was designed to identify frontal lobe dysfunctions of schizophrenic group and to classify into subtypes accordingly. METHODS: Four neuropsychological tests (Wisconsin Card Sorting Test (WCST), Word Fluency Test, Ruff Figural Fluency Test and Grooved Pegboard Test) were administered to 93 schizophrenia or schizophreniform patients diagnosed with the Korean version of SCID. Ten measures (WCST Total Number of Errors, WCST Number of Categories Completed, WCST Conceptual Level Responses, WCST Trials to Complete First Category, WCST Perseverative Responses, GPT Dominant Hand RT, GPT Nondominant Hand RT, Letter Fluency raw score, Category Fluency raw score, RFFT Total Unique Designs) from the four tests were selected by statistical procedure. Latent factors embedded in the frontal lobe function of schizophrenic patients were extracted from the factor analysis, and hierarchical and K-means clustering procedures were used to identify subtypes. To examine the differences among the subtypes, demographic variables, K-WAIS and PANSS were used. RESULTS: (1) The subjects in this study showed significant impairments in the four neuropsychological tests. (2) Through factor analysis, three factors were extracted: Conceptualization, Motor and Fluency. (3) Three cluster solution was considered optimal by cluster analysis. The preserved cluster (n=42) comprised of patients who showed relatively high function in all three factors. This group showed relatively higher function than the other two clusters. However, even the performance of the preserved cluster was 1SD below the norm of the normal people. The conceptualization deficit cluster (n=25) comprised of patients with deficit in conceptualization function. This group was characterized by the clinical symptoms of poor impulse control and active social avoidance suggesting a deficit in the ability to actively organize stimuli utilizing the feedback from the external environment. And finally, the fluency deficit cluster (n=19) showed impairment in fluency. This group was characterized by impairments in the use of abstract-symbolic thinking and the ability to pay attention to relevant stimuli suggesting a severe deficit in the efficiency and flexibility of information withdrawal. These three subtypes didn't differ significantly in age, duration of illness and current dosage of antipsychotics. However, the three groups differed significantly in years of education, IQ and on five items of PANSS. CONCLUSION: This study shows that schizophrenia can be characterized by frontal lobe dysfunctions and divided into three subtypes according to the profile of the frontal lobe dysfunctions. These neurocognitive heterogeneity of schizophrenia, not related to age, duration of illness and dosage of antipsychotic, suggest that different strategies need to be developed in diagnosing and planning rehabilitation programs for schizophrenc patients.