Seventy-seven patients with diagnosis of schizophrenia (62) or schizoaffective disorder (15) were studied 2-20 years since onset of illness, when in a stable condition. The investigation included clinical assessment, measurement of plasma concentrations of neuroleptics and prolactin, computed tomography brain scan, neuropsychological and neurological examination. Outcome of illness was classified according to the presence of chronic psychiatric symptoms and social impairment, and response to neuroleptics according to the effect of treatment in the most recent psychotic episode. Neither outcome nor response to neuroleptics was related to duration of illness. The groups with good and poor outcome differed in premorbid adjustment, age at onset and symptoms of the initial episode, but not in drug bio-availability or prolactin response. Large cerebral ventricles and cognitive impairment, but not neurological 'soft' signs, were associated with unfavourable outcome. The three measures of organicity were not inter-related. No clinical differences were found between chronic patients with and without signs of organic dysfunction. The findings suggest that schizophrenia with good and unfavourable outcome may be separate sub-types. However, the role of organic factors in the latter group remains unclear.

Computer tomography brain scans were carried out on 40 patients with schizophrenia or schizo-affective disorder of 2-20 years duration. Ventricular-brain ratio (VBR) was significantly greater than that of the control group. In six patients the VBR exceeded the control mean + 2 s.d. Among the 13 whose VBR was more than 1 s.d. above the control mean, none had schizo-affective disorder, all but one had chronic illness, and patients with negative symptoms and those with premorbid schizoid traits were over-represented. VBR was unrelated to medical history, age, duration of illness, or neuroleptic treatment. It was not associated with neurological 'soft' signs or cognitive deficit. Among chronic patients, clinical features showed no association with ventricular size. The findings suggest that large ventricles may be related to a sub-type of chronic schizophrenia rather than to its particular clinical features.

No abstract