The sporadic v. familial classification given aetiological heterogeneity: II. Power analyses.

This paper examines the power of the sporadic v. familial method as applied to schizophrenia and major depression. The model used assumes aetiological heterogeneity with a subpopulation of cases due to a 'major' environmental event and the remainder resulting from a generalized single major locus. The findings suggest that, for sample sizes to which it is commonly applied, the sporadic v. familial classification has low power to detect aetiological heterogeneity. When applied to nuclear families, substantial power requires at a minimum 100-150 proband families. If the proportion of environmental cases in the population is low, or the 'test' for environmental aetiology in probands does not have high sensitivity and specificity, the required sample sizes are considerably larger. Adding monozygotic twins increases the power of the method, but including second-degree relatives does not. The optimal approach to the sporadic v. familial method will differ as a function of the frequency of the disorder and the relative effort and expense of examining probands versus family members. Other methods should be considered for discriminating genetic and environmental forms of illness.