Neurocognitive subtypes of chronic schizophrenia.

Performance on four key neurocognitive tasks was used to search for subtypes in 104 DSM-IIIR schizophrenic patients. The tasks were the Wisconsin Card Sorting Test to index executive prefrontal cerebral function, intrusion errors from the California Verbal Learning Test to tap hippocampal-diencephalic mnestic function, bilateral hand performance on the Purdue Pegboard to index fine motor-basal ganglial function, and a pro-rated IQ from the Wechsler Adult Intelligence Scale-Revised to measure general cognitive-cerebral function. Neurocognitive data were analyzed using hierarchical and disjoint clustering procedures with Euclidean distance. A five cluster solution was considered optimal. Cluster 1 (n = 24) comprised patients with selective executive-prefrontal dysfunction; cluster 2 (n = 16) suggested normative function; cluster 3 (n = 20) involved patients with executive-motor/cortico-basal ganglial deficit; cluster 4 (n = 25) suggested dementia/multi-focal disturbance; and cluster 5 (n = 19) consisted of patients with selective motor-basal ganglial deficit. The subtypes differed significantly in age, duration of illness, and extent of hospitalization. Suggestive trends in sex composition and anti-Parkinsonian medication patterns were noted. Neurocognitive tasks combined with cluster analysis have promise in reducing and clarifying the heterogeneity of schizophrenia.