Evidence of Inbreeding Depression on Human Height, a paper with over 1,000 authors! (I exaggerate) It's interesting because it
seems to establish that inbreeding does have a deleterious effect on traits whose genetic architecture is presumably polygenic and additive.
Why is this theoretically important? Because inbreeding depression is often assumed to be driven by the exposure of rare recessive larger effect alleles, which recombine in near relations. Using tens of thousands of individuals from across a dozen European nations the authors found that there is a consistent relationship between inbreeding and reduction in height. As the authors note height is a convenient trait to explore. First, it's highly heritable. 80 to 90 percent of the variation in the population is explained by variation in genes. Second, it's easy to measure. Also, implicit in the paper is the fact that in Europe today there is far less of a environmental effect on height (that's why the heritability value is high). Even in poor European nations most people have enough to eat, so height is highly heritable, allowing for appropriate cross-national comparison. The simplest way to state their results is that all things being equal the offspring of two first cousins will be ~3 cm shorter than those of unrelated individuals. But there are many caveats and qualifications here. First, there are different sources of the depression. Using the most sensitive measure of recent consanguinity, a statistic of run of homozygosity pruned of markers in linkage disequilibrium, the authors did not find a strong effect of magnitude of inbreeding increasing the depression. In other words, it looks like the "bang" sharply diminishes after the first "hit." Second, even genomewide homozygosity has some independent relationship to depression in height. The distinction here is there are genetically homogeneous populations which are nevertheless not inbred. In contrast, there are populations where inbreeding is common, where homozygosity might be lower (an example here might be a Gulf Arab community with lots of African and Persian admixture, increasing the number of heterozygous loci, but where cousin marriage is ubiquitous). The result here suggests that near inbreeding can immediately bring together recessively expressed deleterious alleles and produce a reduction in trait value, but that there is some sort of hit to having a high fraction of homozygous loci as well (I suppose one could posit some sort of genomewide heteryzogote advantage, though I'm skeptical of that). But perhaps the biggest caveat here is population heterogeneity. There is now fair evidence that height differences between European populations are in part genetic. So naturally the magnitude of the decrease due to inbreeding depression is going to vary by the nature of the genetics of height in a given population. Additionally, I'm not quite sure that they've totally accounted for issues of population structure here. Those populations which are naturally very tall may also tend toward greater homozygosity or inbreeding for independent reasons, reducing the effect size of the depression. This is where focusing on Europe is a weakness in this study. I would be very curious about inbreeding depression in Arab populations, for example. The last sentence of their abstract is obvious and intriguing:
Although this exploratory work focuses on height alone, the methodology developed is generally applicable to heritable quantitative traits (QT), paving the way for an investigation into inbreeding effects, and therefore genetic architecture, on a range of QT of biomedical importance.
A major reason people study height is that it is a complex trait which seems to be tractable. You can measure it, and normal variation is relatively wide in realized trait value. The ultimate rationale is often to develop or test methods which can have biomedical application. The implication here is that if inbreeding between first cousins results in a ~3 cm depression in height, who knows what other ailments may be the product of these relationships? There seems a consistent result that offspring of first cousin marriage are less intelligent than similar outbred individuals (again, the magnitude seems to vary by study, but the direction of difference is consistent). But then there's this, The Love That Dare Not Speak Its Surname:
Now a study by the National Society of Genetic Counselors says that having a child with your first cousin raises the risk of a significant birth defect from about 3-to-4 percent to about 4-to-7 percent.
I've seen this study being discussed on "rationalist" websites, illustrating the stupidity of religious taboos on incest. There are two issues which concern me here. First, the aggregate social cost to an increase in congenital defects from 3 to 4 percent is not trivial. Second, the focus on congenital defects ignores the impact on normal human variation. The offspring of cousin marriages may be less healthy, uglier, less intelligent, and perhaps shorter. This doesn't mean that we should ban cousin marriages, anymore than we should ban marriages between stupid ugly people. Rather, it suggests we may need to add some extra parameters to the calculus of the wages of cousin marriage. What about the flip side? In many species there is an equipoise of relatedness (e.g., philopatric frogs). Too close, and inbreeding depression. Too far, and outbreeding depression. Might that be an issue? In the case of height it seems unlikely. This paper indicates that the primary reason for the decrease in height are rare recessive alleles which have a large deleterious effect. The reason for outbreeding depression is most likely going to be some sort of antagonistic epistasis, deleterious gene-gene interactions. I'm skeptical that these are going to be very common (here is a rare example). But it will be an interesting question to address. Looking at long stabilized hybrid populations, such as the Uyghurs and Malagasy might be instructive.