SummaryWe performed a blind family interview study of 226 first-degree relatives of 63 probands meeting DSM-III criteria for schizophrenia, schizoaffective disorder, and bipolar disorder, as diagnosed by the National Institute ot Mental Health Diagnostic Interview Schedule (DIS). A small test-retest reliability study demonstrated good agreement between the proband interviewer and the principal family interviewer for the major diagnostic categories of psychotic disorders. Excellent compliance was obtained, with 85% of living relatives interviewed personally.Three principal findings emerged front the study. First, as expected, bipolar disorder, as defined by DSM-III, displayed a strong familial comportent, comparable to that found by many studies using criteria other than those of DSM-III. Second, patients meeting DSM-III criteria for schizophrenia and schizoaffective disorder displayed a low familial prevalence of schizophrenia. Although initially suprising, this finding is in agreement with the results of several other recent lantily studies of schizophrenia. Upon comparing our results with those of other recent family studies of schizophrenia, it appears that the familial component in schizophrenia tnay be less than was estimated by earlier studies using older and “broader” definitions of schizophrenia.Third, we found that patients meeting DSM-III criteria for schizophrenia appeared genetically heterogeneous. Those who had displayed a superimposed full affective syndrome at some tinte in the course of their illness, together with those probands meeting DSM-III criteria for schizoaffective disorder, displayed a high familial prevalence of major affective disorder, similar to that found in the families of the bipolar probands. On the other hand, “pure” DSM-III schizophrenie probands, who had never experienced a superimposed full affective syndrome, displayed a low familial prevalence of major affective disorder, similar to that found in the general population. These findings favor the possibility that probands meeting DSM-III criteria for schizophrenia, but displaying a superimposed full affective syndrome, may in sonie cases have a disorder genetically relatcd to major affective disorder.Further prospective family interview studies, using DSM-III criteria and larger samples, will be necessary to test these preliminary impressions.
A scalable method for the generation of small test sets