The Prognostic Importance of Genetic Factors in Functional Psychoses

¹ Gaustad Hospital, Oslo, Norway

In two follow-up series of first admissions of functional psychoses to Gaustad Hospital we have 972 cases admitted between 1938 and 1950, and 706 cases admitted between 1951 and 1957. For these two series the genetic background has been analysed by comparing the clinical pictures of the probands with the clinical pictures of the relatives, evaluated from their case histories. As a rule the relatives tend to have clinical pictures similar to those of the probands, although there are many exceptions.

Utilizing computer techniques, prognostic models predicting the risk of schizophrenic defects were constructed based on coded clinical symptoms, social and genetic variables. The prognostic weights of the social variables varied considerably over the two time periods, which reduces the general applicability of their inclusion in a prognostic model. For the total sample of 1938-57 first admissions a model was constructed, containing five of the most predictive clinical items along with the genetic variable of presence of schizophrenic psychoses among siblings, parents, grandparents, uncles or aunts. The genetic variable turned out to be the fourth best predictor. Accordingly, we consider it well justified to include the genetic variable in a prognostic model. The model gives about 80 per cent. correct prediction of long term outcome.