To the Editor: Schizophrenia, a disease with a strong genetic component, has not disappeared, despite the fact that affected patients have lower fertility than the general population. Jari Haukka, Ph.D., et al. (1) tried to explain this apparent paradox by testing the hypothesis that the relatives of schizophrenia patients have higher fertility than the general population. Not surprisingly, the study did not confirm this hypothesis.

The transmission of the schizophrenia phenotype does not conform to a simple Mendelian pattern, even when we allow for incomplete penetrance. There are two models of transmission that fit the data available in the literature: a polygenic model and a multifactorial model (2). In a polygenic model, liability to develop a disease is continuously distributed in the population because of the additive effects of multiple genes at different loci. Only the individuals whose liability exceeds a certain threshold will manifest the disorder. The relatives of affected individuals have an increased mean liability compared with the population as a whole, resulting in more relatives manifesting the disorder. Multifactorial models allow extension of this concept, such that liability can be contributed by genetic and environmental factors in an additive fashion. Most of the discussion regarding the polygenic model applies to the latter.

Given the high prevalence of schizophrenia, these models suggest that the genes that give liability to the illness are not confined to schizophrenia patients and their relatives but are present in variable quantities in all humans, with a Gaussian distribution. Looking at the relatives of schizophrenia patients for a fertility advantage to explain the persistence of these genes is therefore not the right approach. These genes are an intricate part of the nature of Homo sapiens. They are not just an advantage to some but indispensable to all of us. The price paid is that 1% of the population develops schizophrenia.

When we look at diabetes as an analogy, it exists because evolution has developed a system dependent on insulin for the metabolism of glucose. The advantages offered by such a mechanism outweigh the disadvantages of having a proportion of the population affected by diabetes. The reason that the genes that give liability to diabetes do not disappear from the human genetic endowment, despite lower-than-average fertility in the individuals affected (3), is therefore clear when we look at the human population in its totality but would defy us if we analyzed only the fertility of siblings of diabetic patients. The same is true for schizophrenia.