Ex uno multi: subtyping the schizophrenic syndrome.

We studied 93 chronic schizophrenic inpatients, who met the Research Diagnostic Criteria for schizophrenia. Data on a number of historical, epidemiologic, phenomenologic, biochemical, neuropathological, and treatment-response variables were analyzed, using univariate and multivariate statistical analyses. Patients were classified into pairs of subgroups, according to each of the following seven dimensions: (1) ventricle/brain ratio (VBR) assessed on computed tomography scans (normal vs. abnormal); (2) premorbid adjustment (good vs. poor); (3) therapeutic response to neuroleptics (good vs. poor); (4) platelet monoamine oxidase (MAO) activity (low vs. high); (5) paranoid features (present vs. absent); (6) tardive dyskinesia (present vs. absence); and (7) hemispheric asymmetry on computed tomography scans (normal vs. abnormal). These seven dimensions were chosen because earlier studies had shown that the variables involved were operationally definable and were of potential relevance to the subgrouping of schizophrenic patients. Our results suggested that two biological variables, viz., VBR and platelet MAO activity, might be useful in identifying two rather distinct subgroups among chronic schizophrenic patients. A subgroup with large VBR was associated with poor premorbid adjustment, neurological impairment, and poor therapeutic response to neuroleptics, while the subgroup with low platelet MAO activity was characterized by the presence of paranoid features and tardive dyskinesia. Possible explanations, implications, and limitations of our findings are discussed.