Human Nature: Science, Technology, and Life.

Incest and Eugenics

Does science support our laws against incest and cousin marriage? If so, does it also support other laws that would restrict sexual or procreative freedom in the name of genetic health?

To longtime readers of Human Nature, this question should be, if you'll pardon the term, familiar. A few years ago, we looked at the science and ethics of "The Love That Dare Not Speak Its Surname." Then we examined the prevalence of inbreeding in nature. Then we considered the awkward question of why, if incest is too genetically risky to permit, maternity in your 40s isn't.

Now biologist Hamish Spencer and political scientist Diane Paul, writing in PLOS Biology, have reviewed the history of U.S. laws against cousin marriage, along with their scientific basis. And again, the evidence raises unsettling implications.

They start with the statistical case against restricting cousin marriages:

[T]he National Society of Genetic Counselors (NSGC) convened a group of experts to review existing studies on risks to offspring and issue recommendations for clinical practice. Their report concluded that the risks of a first-cousin union were generally much smaller than assumed—about 1.7%-2% above the background risk for congenital defects and 4.4% for pre-reproductive mortality—and did not warrant any special preconception testing. In the authors' view, neither the stigma that attaches to such unions in North America nor the laws that bar them were scientifically well-grounded.

But Paul and Spencer point out that the data aren't clear-cut. First, "statistics on the risks associated with cousin marriage are necessarily averages across many traits, and they are likely to be different for different populations." And second, it's

inappropriate to extrapolate findings from largely outbred populations with occasional first-cousin marriages to populations with high coefficients of inbreeding and vice-versa. Standard calculations, such as the commonly cited 3% additional risk, examine a pedigree in which the ancestors (usually grandparents) are assumed to be unrelated. In North America, marriages between consanguineal kin are strongly discouraged. But such an assumption is unwarranted in the case of UK Pakistanis, who have emigrated from a country where such marriage is traditional and for whom it is estimated that roughly 55%-59% of marriages continue to be between first cousins. Thus, the usual risk estimates are misleading: data from the English West Midlands suggest that British Pakistanis account for only ~4.1% of births, but about 33% of the autosomal recessive metabolic errors recorded at birth.

In other words, the American calculations understate the risk for an already inbred population such as British Pakistanis. And calculations based on British Pakistanis overstate the risk for most American cousin couples. You can't draw a uniform line against cousin marriages based on science. Arguably, for the same reasons, you can't draw a uniform line against sibling incest.

The same case can be made against a uniform age of sexual consent. The authors point out,

Beginning in the 1860s, many states passed anti-miscegenation laws, increased the statutory age of marriage, and adopted or expanded medical and mental-capacity restrictions in marriage law. Thus, laws prohibiting cousin marriage were but one aspect of a more general trend to broaden state authority in areas previously considered private.

As Human Nature has noted before, the age of actual maturity varies considerably depending on the person and the type of maturity (sexual, cognitive, emotional) involved. Granted, lawmakers have to draw lines somewhere. But let's not pretend such consistent lines are consistently apt.

Moreover, Paul and Spencer raise a far more troubling problem: The increase in genetic risk caused by cousin marriages among British Pakistanis may actually be overstated, for a curious reason.

[F]or a variety of reasons (including fear that a cousin marriage would result in their being blamed for any birth defects), UK Pakistanis are less likely to use prenatal testing and to terminate pregnancies. Thus the population attributable risk of genetic diseases at birth due to inbreeding may be skewed by prenatal elimination of affected fetuses in non-inbred populations.

In other words, many of the birth defects cited by British politicians as grounds for restricting cousin marriages may actually be the result not of cousin marriage, but of failure to screen and abort defective fetuses. So, in addition to maternity in your 40s, we now have a second logical target for genetic regulation: If inbreeding is too dangerous, what about "inflicting" maladies on your children by failing to screen the embryos? If you know you carry bad genes—and particularly if you're at higher risk of passing down a serious disease than most sibling couples would be—shouldn't we police your procreation just as carefully?

Incest and Delayed Motherhood

On Friday I wrote about homosexuality, polygamy, and incest. The gist of the piece was that our categorical bans on these practices are losing their justification. As societies embrace privacy, the acceptable basis for restricting sexual behavior has been reduced to harm. And the evidence that these practices are harmful is weak.

In the case of incest, I looked at the scientific objection -- inbreeding -- at a level one step removed. That is, not in brother-sister coupling, but in cousin marriage. This is an emerging controversy in Britain, thanks largely to immigration from Pakistan, where the practice is common.

Over the weekend, two more articles on this topic appeared in the British press. Let's take a look at them.

First, an estimate of the scale of the practice. "Over a billion people worldwide live in regions where 20%-50% of marriages are consanguineous -- that is where the partners are descended from the same ancestor," reports Emma Wilkinson of BBC News. In Britain, Wilkinson cites an unfolding study in Bradford, where half the kids are from Pakistani parents. A pediatrician at the local teaching hospital reports that 70 percent of the first 1,100 Pakistani women recruited for the study are offspring of consanguineous marriages.

Second, some political background, courtesy of Ian Sample in the Guardian. Three years ago, a member of parliament from the left-leaning Labour Party was denounced for suggesting that cousin marriage should be discouraged as a genetically harmful practice. "We have campaigns about the health effects of drinking, smoking and overeating," the MP pointed out. Why not mount a similar information campaign about cousin marriage? A few months ago, a second MP echoed this argument and was rebuked by the prime minister's office.

The actual risk-multiplication effect of cousin marriage isn't clear. A study I cited six years ago concluded that having a child with your first cousin increased the risk of a significant birth defect from about 3-to-4 percent to about 4-to-7 percent. Wilkinson cites data showing that "since 1997 there have been 902 British children born with neurodegenerative conditions and 8% of those were in Bradford which only has 1% of the population." This appears to be the basis for Sample's report that "rare inherited brain disorders are eight times higher among Pakistani children born to married cousins than those born to unrelated parents." But Wilkinson adds that Australian geneticist Alan Bittles, supposedly the top expert on this subject,

has collated data on infant mortality in children born within first-cousin marriages from around the world and found that the extra increased risk of death is 1.2%. In terms of birth defects, he says, the risks rise from about 2% in the general population to 4% when the parents are closely related.

If Bittles' numbers are correct, they substantiate a somewhat embarrassing point made by defenders of cousin marriage. Embarrassing, that is, to all of us good Western folk who turn up our noses at the practice. The British Down's Syndrome Association has posted a chart showing the risk of producing a baby with the syndrome at various maternal ages. From age 20 to age 31, the risk doubles. From 31 to 35, it doubles again. From 35 to 38, it doubles again. From 38 to 41, it more than doubles again. Each delay multiplies the risk as much as cousin marriage multiplies the risks of all birth defects combined. By age 45, the probability of Down syndrome alone roughly matches the 4 percent cumulative risk of birth defects from cousin marriage.

Which brings us to the Elizabeth Edwards question. As Suz Redfearn reported in Slate four years ago, Edwards gave birth to her two youngest children, Emma Claire and Jack, when she was 48 and 50. Redfearn thinks Edwards used donor eggs. Edwards won't say. If Edwards used her own eggs, the Down syndrome chart puts her probability of the disease at 1 in 11 for Emma Claire and 1 in 6 for Jack. That's two to four times the risk of any birth defect from cousin marriage.

Should women be allowed to have babies well into their 40s? If so, how can you justify restrictions on cousin marriage? For that matter, what about sibling incest? Theoretically, given a pool of recessive disease genes, reproducing with a sibling instead of a first cousin quadruples the risk of defective offspring. This probably overstates the actual effect, since population studies don't show quadrupling as degrees of consanguinity increase. But even if the birth-defect rate is a worst-case 17 percent, that's no higher than the risk of Down syndrome at the age when Elizabeth Edwards had her fourth child.

For what it's worth, it looks as though Britain may take a middle course: no legal restrictions on cousin marriage, but no indifference, either. Bittles and others are proposing to reduce birth defects through counseling, genetic screening, and public education in communities that practice cousin marriage.

My guess is that this is how governments will manage unconventional sex practices in the next century. We can't stop people from doing what they want to do. We'll tell them what's generally dangerous. And if they can adequately reduce the medical risks, by wearing a condom or taking a genetic test, we'll look the other way. We'll speak the language of science, or none at all.