Location of the adventitious factors in haemolytic disease of the newborn.

It is by now well established that the frequencies of ABO mating types among parents of children with Rh haemolytic disease of the newborn are not the same as in the general population. There is a reduced frequency of matings in which the mother has antibodies to ABO-antigens possessed by her husband. It is possible that incompatible foetal blood escaping into the mother's circulation is so rapidly destroyed that its ability either to sensitize or later to immunize the mother against Rh substance is impaired. Nevanlinna and Vainio (1956) have also shown that within affected families the proportion of healthy children who are incompatible withthemother for the ABO groups is even lower than the proportion of affected ones. They have argued that because, on average, healthy children occur earlier in these families than do affected ones, the protective effect of ABO-incompatibility is chiefly exercised at the stage of sensitization (sensibilization in their terms) rather than during the pregnancy when antibodies are first produced in sufficient quantities to cause disease. Knox and Walker (1957) constructed an argument based upon the birth rank distribution of affected infants which led to a conclusion in the form of two alternatives. One was complex, requiring an extra genetic determinant, together with a risk of sensitization of the order of 0 39 at each Rh-positive pregnancy, and a similar risk of immunization at each subsequent Rh-positive pregnancy. The other was much simpler (and therefore preferred), requiring no important extra genetic determinants and a single important adventitious factor related to only one of the pregnancies, with a risk of occurrence of 1/17. The method could not distinguish which pregnancy this was. It has since occurred to us that the protection conferred by ABO-incompatibility, apart from its intrinsic interest, may also be used as a tool to extend our previous argument. For if it can be shown precisely at which pregnancy or pregnancies the protection is conferred, then that also must be the location of an important adventitious factor against which the incompatibility protects. It is therefore necessary to press the arguments of Nevanlinna and Vainio to a more precise and critical stage; to deduce in detail the expected consequences of the hypothesis they have set up and of alternative hypotheses; to examine the facts in relationship to these deductions. This is the purpose of the present paper.