The Dating Advantage Theory
A new study argues that the same genes the increase schizophrenia risk also make you a more attractive person to date and mate. For example, those genes, when expressed in the general population at levels insufficient to lead to clinically diagnosed schizophrenia may enhance courtship related traits such as "creativity, emotional sensitivity, and expressiveness."
(The study is Del Giudice M (2010) Reduced Fertility in Patients' Families Is Consistent with the Sexual Selection Model of Schizophrenia and Schizotypy. PLoS ONE 5(12): e16040. doi:10.1371/journal.pone.0016040).
As the article explains, one theory is that:
Schizotypy [i.e. genes that predispose a person to schizophrenia] . . . act[s] as an “amplifier” of individual differences in genetic fitness and condition: high-schizotypy individuals would be more likely to achieve outstanding mating success when they have high genetic fitness and/or grow up in good environmental conditions, but also more likely to develop schizophrenia (and to suffer reduced mating success) when they have low genetic fitness and/or grow up in poor conditions. Thus, schizotypy itself can be sexually selected, possibly more strongly so in populations characterized by high levels of mating competition and short-term mating patterns, where displays of good genetic quality are more critical to successful courtship [than traits associated with success in a long term committed couple relationship].
In other words, schizophrenia related genes may be particularly adaptive in societies with "love marriage" as opposed to arranged marriage, because traits that pre-dispose you to schizophrenia also make you someone who is fun to date. The fact that schizophrenia generally manifests clearly only towards the end of puberty or early young adulthood, also suggests that full blown schizophrenia may often be invisible until after marriage in societies that do not, as modern American societies do, favor historically late ages at first marriage. In other words, if a typical couple gets engaged in their late teens, the most negative aspects of schitzotypy are more likely to be invisible.
This analysis generalizes a similar evolutionary explanation for the persistence of bipolar disorder, which presents similar evolutionary issues to schizophrenia, and is also very common in individuals with extraordinary creative output, a trait that is considered fitness enhancing, at least in modest frequencies, in a society. There is some anecdotal evidence points out, for example, in the graphic novel history of modern mathematics found in Logicomix that there may be a similar association between schizophrenia and extraordinary output in mathematics and logic.
The Undercounted Descendants Theory
The study also hypothesizes that schizophrenics, particularly men, may be more likely to have unacknowledged children outside of marriage, artificially suppressing their measured fertility rates: "This is not an unlikely possibility, given that schizotypal personality traits specifically predict increased engagement in short-term mating and casual sex." This hypothesis, which is powerful in traditional societies, may have less force in societies where contraception is widely used and effective in these kinds of relationships.
The greater fertility impact in men than in women, however, may also be explained by the fact that schizophrenia is more likely to manifest before men have had children (because men have an earlier onset on average, and later peak reproductive years than women) than it is to manifest before women have had children.
Limits Of The Methodology
The study is primarily simply an exposition of a mathematical model that shows that the explanation proposed would be sufficient to explain the survival of genes associated with schizophrenia with mating advantage effects that are within a range that is plausible, without actually measuring if those advantages really exist. But, it has considerable force, despite its modest empirical grounding, because it is hard to find another explanation that could explain the facts that are observed. If the model proposed in this paper is wrong, one has to find some other model of how a strongly hereditary trait that is strongly selected against in many respects has remained stable in population frequency.
Persistent, fitness reducing genetic traits very frequently have an aspect that is fitness enhancing, or was at some time in the past. The sickle cell anemia gene, which is also connected to malaria resistance is the most commonly cited example of this kind of trait, but many strongly hereditary fitness reducing conditions share that feature. A more recent example (details behind pay wall) that is largely speculative at this point, is the hypothesis that a key breast cancer risk gene (BRCA1) may also enhance intelligence, and that the cancer risks and cognitive benefits may both relate to the production of connective tissue proteins. The problem with this model in the case of schizophrenia, however, is that almost all of the genetic conditions of this type known so far have simple, single gene Mendelian inheritance patterns, something that is not what is behind the strongly hereditary component of schizophrenia. Extensive research directed at the genetics of schizophrenia have pretty much ruled out such a simple cause, and it is much harder to develop a "some mutations good, many mutations bad" scenario in a more complicated genetic model. I am not aware of any hereditary condition that has been definitively established as having that kind of phenotype structure.
Breast cancer risk genes, like schizophrenia risk genes, are also notable in manifesting relatively late in life, which reduces their selective impact. "Schizophrenia occurs equally in males and females, although typically appears earlier in men—the peak ages of onset are 20–28 years for males and 26–32 years for females. Onset in childhood is much rarer, as is onset in middle- or old age."
Keep in mind that life expectancy in the Roman Empire (and much of the period before it and after it) was twenty to thirty years, and that historically people had children young, so conditions that first have negative fitness effects when you are in your later teens or early twenties, or later, have had only slight selective effects until the last dozen or so generations or less, in the West, and for half a dozen generations or so in much of the rest of the world. So both breast cancer risk genes and schizophrenia risk genes may have had far less of an impact of evolutionary fitness for 99%+ of their existence in the human genome.
Probably the weakest piece of the model is the significance attached to environmental impacts on the likelihood that schizophrenia will manifest, which seems to be quite low empirically.
A Mutation Model
Perhaps the best alternative explanation to those suggested in the study would be a model similar to that advanced for autism (where that evidence seems to suggest that a majority of all cases may be first generation mutations attributable to advanced parental age; many kinds of mental retardation are also due primarily to "first generation" mutations), which sees the condition as being a product of many different kinds of common mutations that all have the same phenotypic effect (perhaps because they all interrupt the same delicate neurological system). Thus, there may be enough "first generation" schizophrenics to make up for a deficit in schizophrenic children of schizophrenics. The main pieces of evidence supporting this as part of the explanation are the facts that schizophrenia, like autism, is associated with advanced paternal age probably due to the increased frequency of new mutations in children of older fathers, and that both schizophrenia and bipolar disorder and autism seem to be associated more closely with the total number of atypical mutations (or subtypes of atypical mutations) in a person's genome than with any one mutation in a particular disease that accounts for a large share of all cases individually.
Schizophrenia, bipolar disorder and autism may all be the "fevers" of mental health. In other words, they may be common observable symptoms because a multitude of underlying causes all disrupt the delicate system that causes each symptom (schizophrenia, bipolar disorder, autism and fever respectively) to manifest.
As long as mutation rates among parents in the general population are high enough to make up for deficits in fertility in schizophrenics (which may be smaller than reported due to unacknowledged children of schizophrenic men who have lower fertility rates than schizophrenic women), schizophrenia can persist in the population at a steady prevalence despite being selected against.
The Theories Are Mutually Compatible And Avenues For Further Investigation
Of course, all three of these theories are compatible with each other. The larger the proportion of "first generation schizophrenics" (something that should be possible to estimate with reasonable accuracy from the strength of the paternal age-schizophrenia relationship), and the larger the number of unacknowledged children that men with schizophrenia are having (which might be best estimated as a percentage in excess of the percentage of unacknowledged children in the general population for which there is better empirical data with the percentage estimated by surveyed extramarital sex frequency of schizophrenic men relative to that of a control group), the less pressure there is on courtship trait advantages conferred by subclinical schitzotypy to have to be large and on subclinicial schitzotypy to be common in the general population (a frequency that can be estimated from the share of diagnosed schizophrenics with a family history of schizophrenia and that is, given the literature, probably quite low but non-zero) to explain the condition's evolutionary persistence.
Similarly, if it takes multiple mutations to create symptomatic schizophrenia, and a significant share of schitzotypy genes are "first generation", it is quite plausible that there is a significant rate of subclinical schitzotypy in the general population, which produces descendants who inherit a combination of schitzotypy genes that produce clinical schitzophrenia. One way to test this hypothesis would be to look not merely at advanced parental age as a risk factor, but also to examine advanced male grandparental age as a risk factor. A new mutation accumulation model with a threshold number of mutations required for an individual to be symptomatic would be supported and could be quantified at an order of magnitude level to the extent that advanced male grandparental age was a factor independent of advanced parental age. A study of advanced male grandparental age, of course, has the virtue of being something that can be cheaply, unobtrusively, and accurately determined from participants in existing studies.
The "courtship advantage" model proposed in this study for schitzotypy genes, like the studies showing creativity advantages for sufferers of bipolar disorders, and analysis showing benefits in modern society from traits associated with the autism spectrum disorder all point to the importance of a neurodiversity model for thinking about human hereditiary cognitive differences.
If human hereditary cognitive differences are classified as flaws or diseases on one hand, and as enhancements on the other, there is a strong incentive for society and individual decision-makers to use eugenic approaches and gene therapy to wipe out the flaws from the human genome. But, if we recognize that some of the flaws are inseparable from cognitive traits that have value at some frequency (even if it is a low frequency) in our society, then efforts to wipe out genetic "flaws" in the population in the interest of improving mankind should be viewed with far greater skepticism.