Editorial: Genetic Epidemiology
Genetic epidemiology arose from the field of population genetics. Fisher, Haldane, and Wright set out the mathematical underpinnings of modern population genetics and a vigorous branch of genetics developed. As long as human and medical genetics was based on phenotypes far removed from gene action, most approaches had to rely on various mathematical-statistical considerations. Classic population genetics with its subfield of formal genetics provided many of the tools for such work. The development of human biochemical genetics, human somatic cell genetics, and human cytogenetics shifted the action from the "calculator" to the laboratory. Abnormal gene products and abnormal chromosome sets could easily be recognized, and the majority of human and medical geneticists shifted their attention to work in these areas. However, it was soon realized that many common diseases were familial and could not readily be approached by unifactorial models of Mendelian transmission or by search for cytogenetic abnormalities. The operation of genetic, environmental, and cultural factors common to families was often suggested in these diseases. Usually the nature of the genetic or environmental agents remained unknown so that laboratory study of specific genetic and environmental agents was impossible. Here is where the newly established field of genetic epidemiology came in. The development of computers made complex calculations and the handling of large bodies of data relatively easy and allowed testing of a variety of genetical models. Furthermore, the desire of many researchers to work on "relevant" rather than theoretical topics together with the interest of granting agencies to support disease-related work gave the field a significant boost. Soon, a small group of investigators took data from several common diseases with familial aggregation and applied new genetic-statistical methodologies to the findings. The principal conceptual difference between classic epidemiology and genetic epidemiology was the attempt of genetic epidemiologists to detect subgroups of individuals who because of their genotypes were more susceptible to develop certain diseases. When biological methods could be applied the data were often clearcut and noncontroversial. Some examples include the role of the abnormal LDL receptor in a subset of patients with hypercholesterolemia and the role of the HLA-D 3 and 4 alleles in susceptibility to insulin-dependent diabetes. Attempts to fit genetic models to diseases with phenotypes far removed from gene action were less satisfactory since a variety of genetic models fitted the data. Schizophrenia is an example. Presumably, in the absence of biologic leads other than the clinical diagnosis alone, genetic heterogeneity and other confounding factors made meaningful answers impossible. The most elegant and ingenious statistics cannot substitute for lack of biologic insight! 0 1984 Alan R. Liss, Inc.