GENETIC MODELS OF
Raymond R. Crowe
WHEN the long-standing "nature versus nurture" controversy in psychiatry was finally resolved by the adoption studies, demonstrating beyond a reasonable doubt the genetic predisposition to the major mental disorders, attention finally turned to the more productive question of how both genetic and environmental factors are involved in transmitting disease; that is, transmission models.
The earliest of these transmission models were Mendel's laws of inheritance, and early workers in psychiatric genetics naturally attempted to apply these models to the diseases they were studying. However, unlike many medical diseases that fit the Mendelian ratios with such precision that another explanation would be hard to imagine, none of the psychiatric disorders gave an acceptable fit. This led to models modifying Mendel's laws by postulating such variables as incomplete penetrance and additional genes. These explanations were never widely accepted, probably because enough modifications could make the model fit almost any data set and, moreover, no means existed for statistically testing the fit with a probability level.
A major development occurred in the 1960s with the adaptation of the laws of quantitative genetics to the study of discrete traits; that is, the multifactorial threshold model. In the 1970s, this model was expanded to incorporate multiple disease forms, sex differences, and most important, it was rendered statistically testable for goodness of fit. Meanwhile, sophisticated single-locus models that lent themselves to statistical testing were also developed. Finally, the increasing availability of computers made possible analyses that would have been impossible in the early days of model construction.
The term "transmission model" is preferred to "genetic model" because the models do not necessarily require an assumption that the disorder is genetic. Analysis by trans