The nosology of schizoaffective psychosis.

The independent status of schizoaffective psychosis is reviewed in relation to 6 hypotheses, using pattern of inheritance, treatment response, course and outcome, and platelet 5-HT uptake as discriminating variables. The 'coincidence of 2 diseases' hypothesis would predict an annual frequency of some 2 per 10(8), compared with the observed 2 per 10(5). The 'variant of schizophrenia' hypothesis has been supported from patient groupings weighted in favour of chronicity, but little else. The 'variant of affective psychosis' hypothesis has been supported in the manic sub-group by evidence of lithium response, and outcome better than schizophrenia (but worse than mania). However in the depressed sub-group, despite responsiveness to ECT, a tendency towards chronicity is observed in a substantial proportion. The 'provisional diagnosis' hypothesis is supported by heterogeneity of outcome, and the presence of both schizophrenia and affective disorder in families of schizoaffective patients, prompting a distinction between 'mainly schizophrenic' and 'mainly affective' groupings. Evidence for the 'third psychosis' hypothesis is mainly genetic, suggesting a small aetiologically distinct subgroup which breeds true. The authors conclude with an 'interacting process' hypothesis where distinct diseases can cause, or interact to cause, the same symptoms via some final common pathway of psychological dysfunction.