Only authorized persons admitted: the quantitative genetics of health and disease.

The scope of quantitation in genetics ranges from traits that are defined by quantity to those that can be quantitated not at all, or only with difficulty and on artificial scales of specious interpretability. Since disease is a complex process, there is a serious risk that a projection of it on an arbitrary system of measurements may be insensitive or frankly misleading. A hypothetical but plausible example (the "brittle model") of a purely genetic trait (but one with zero heritability) is presented. In addition an illustration is given that competition in pathways to an endpoint (such as death) may lead to a counterintuitive reversal of the positive skewness in the component processes (the "bingo model"). These paradoxical results reflect the perils of inadequate descriptors and an irrational reliance on stock methods of looking at the inheritance of disease. Well-known studies by Brown and Goldstein, Knudson, Paigen. Armitage and Doll, and Moolgavkar are cited as examples of how medical geneticists might rationally approach the genetics of common and complicated diseases.