The inheritance of liability to certain diseases, estimated from the incidence among relatives

It is now commonly recognized that many diseases that are not inherited in a simple manner have, nevertheless, some hereditary basis. The evidence that heredity plays some part comes from the observation that the incidence of the disease is higher among the relatives of affected individuals than it is in the general population. An increased incidence among relatives does not, however, go far toward providing an amwer to the important question of how strong the hereditary factor is, because the difference of incidence has no simple genetic interpretation. The relative importance of heredity and environment in such a case is clearly a problem of quantitative genetics. The usual methods of quantitative genetics, however, are not immediately applicable because these are based on correlations between relatives in respect of some ‘graded’ character measurable on a continuous scale Data in the form of incidences refer, in contrast, to an ‘all-or-none’ classification; individuals either have the disease or they do not. Though the affected individuals may sometimes be graded according to the degree of severity of their symptoms, the normal individuals, who are the majority, cannot be graded by the degree of their normality. The purpose of this paper is to suggest that the method developed in quantitative genetics for dealing with ‘threshold characters ’ is applicable to data on the incidence of diseases, and that by its use we can get further towards an answer to the question of the relative importance of heredity and environment. (A fuller account of the method as applied in quantitative genetics will be found in Falconer, 1960.) The question of most general interest about the genetic causation of a disease that is not simply inherited is probably the relative importance of heredity as a causative agent. This questioii is meaningful only when stated in terms of amounts of variation; i.e. the variation between individuals that causes some to be affected and some not. What fraction of this variation is attributable to genetic differences between individuals? This fraction may be called the ‘degree of genetic determination ’. Unfortumtely the degree of genetic determination cannot be estimated from human data, unless possibly by the use of twins, but a related quantity, the heritability ’, can be estimated. The distinction between the degree of genetic determination and the heritability is as follows. Two kinds of genetic variation have to be distinguished, ‘additive’ and ‘non-additive’. The additiw genetic variance is attributable to the average effects of genes considered singly, as transmitted in the gametes. The non-additive genetic variance is attributable to the additional effects of these genes when combined in diploid genotypes. It therefore arises from dominance and interaction between genes at different loci ; if there is no dominance or interaction there can be no non-additive variance. The degree of