The prenatal wages of interracial relationships

Photo credit: Robert Hanashiro, USA TODAY

It is the norm today to discuss race as a social construct. Less fashionable is it to explore race as a biological concept. When there's no up or downside and the discussion is abstract I think most people can get away with benign neglect in regards to the second; but when your health is at issue people's ears perk up. The HapMap. Here's the first paragraph in Wikipedia on the HapMap:

The International HapMap Project is an organization whose goal is to develop a haplotype map of the human genome (the HapMap), which will describe the common patterns of human genetic variation. The HapMap is expected to be a key resource for researchers to find genetic variants affecting health, disease and responses to drugs and environmental factors....

The rationale behind the HapMap is practical, but its yield has a great deal of basic science relevance. Last year's story about human adaptive acceleration does have connections with disease, but fundamentally its goal was fixated on questions of evolutionary anthropology and demographics, not clinical health. Of course "human genetic variation" can mean many things. There are after all de novo mutations which arise across the generations so that each child differs from their parents in a subset of deleterious mutants. This isn't the focus of something like the HapMap (note the point about "common patterns"), but at some point I assume that full genomic sequencing will make these comparisons more concrete so as to help develop clinical regimes that are extremely individual tailored.

But on to the main topic at hand, next month in the American Journal of Obstetrics and Gynecology there will be an article on prenatal risks to pregnancies of Asian-white couples. From EurekaAlert, Asian-white couples face distinct pregnancy risks...:

It's difficult to estimate the prevalence of Asian-white couples, but 14.3 percent of Americans reporting Asian race in the U.S. Census Bureau's 2000 survey also reported being of mixed Asian-white ancestry. Although past studies have looked at ethnic differences in perinatal outcomes, the majority of research has focused on white- African-American couples. Few studies have focused specifically on Asian-white couples, said El-Sayed, who is also associate chief of maternal-fetal medicine.


More specifically, the researchers found that white mother/Asian father couples had the lowest rate (23 percent) of caesarean delivery, while Asian mother/white father couples had the highest rate (33.2 percent). Because birth weights between these two groups were similar, the researchers say the findings suggest that the average Asian woman's pelvis may be smaller than the average white woman's and less able to accommodate babies of a certain size. (Asian couples had babies with the lowest median birth weight, so caesarean delivery was less common among those women.)

It's important for clinicians to know which women may have an increased risk of caesarean delivery, so they can conduct proper counseling prior to childbirth, El-Sayed said.

El-Sayed and his colleagues also found that the incidence of gestational diabetes was lowest among white couples at 1.61 percent and highest among Asian couples at 5.73 percent - and just under 4 percent for Asian-white couples. These findings weren't altogether surprising: past studies have shown an increased risk of diabetes among Asian couples, which researchers attribute to an underlying genetic predisposition. But the interesting finding, El-Sayed said, was that the risk for interracial couples was about the same regardless of which parent was Asian.

If you read this weblog you won't be too surprised. Two years ago there was a paper in Nature Genetics which reported a result that African Americans may have heightened risks for heart problems because they are a genetically admixed population. It seems that an allele which African Americans inherited from their European ancestors (20-25% of African American ancestry is European) results in inflammatory problems against a West African genetic background (deleterious gene-gene interactions). Racially mixed individuals in Hawaii seem to be more likely to be obese; the process might be purely cultural in derivation, but there might also be risks due to interaction of metabolically related loci. Finally, different frequencies in blood groups can result in higher miscarriage rates for some combinations.

The first example is one where Europeans and West Africans are disjoint in allelic frequency; the genetic variation among Europeans is just not found in West Africa. The last example is one where the groups overlap, but, the average between group differences matter. In both of these cases there isn't something mystical going on; God is not smiting those who are sinning by crossing racial lines. Human spontaneous abortion rates are high. Much of this might be due to mother-child immune system responses. It shouldn't surprise anyone that people from very genetically distinct populations have very different immune profiles. One byproduct of this is to produce new and novel genetic combinations in offspring. This might result in a more robust immune system in terms of defenses against common diseases since the immune profile is so rare that pathogens haven't had to adapt against it. But, it also causes problems when it comes to tissue matching for organ transplants; mixed-race children exhibit a synthetic immune profile which has similarities with, but is rarely identical to, individuals from the two populations which their parents derive from. All this is presupposed on the viability of the offspring. What about the large number of spontaneous abortions (miscarriages)? There is some data which might suggest that genetic relatedness increases reproductive fitness, possibly because of reduced risk of immune incompatibilities between mother and fetus when the father is more closely related to the mother. It stands to reason then that as the father becomes more genetically distant the likelihood of incompatibilities might increase.

All of this means that genes matter, and they matter in ways we can possibly predict. For example, imagine a population where everyone has blood group O (many Native American tribes are like this) and another where everyone has blood A (I believe this is close to true for the Blackfeet). If you make the claim (which is empirically defensible, see links above) that women who are O blood group will have much higher rates of miscarriage when they are carrying babies from males who are of blood group A, you can conclude that matings between the first population and second will have reduced fertility. But, the same applies to people from populations where there is a fair amount of blood group variation. If a woman from a village in Norway of blood group O marries a man from the same village of blood group A, even if they are cousins she will have the same heightened risk of miscarriage. There's nothing mystical in the blood here, it's simple genetics. The likelihood is a product of clear and distinct factors. Between population differences are often simply exaggerations of within population difference, though not always. Just because on any random locus 85% of the variation is found within populations and 15% between, that does not mean that on all loci 85% of the variation is found within populations and 15% between. On some genes nearly all the variance is between population and almost none of it within population.

Because of the results on Caesarean section rates they adduce that there is a pelvic size difference between Asian women and white women. Objective male observer acquaintances of mine have generally tended to back up this phenotypic difference between the populations. But, that does not necessarily speak to individual experiences, that is, if the authors corrected for pelvic size on an individual level they might not need to make recourse to population level generalizations. The results on gestational diabetes not being effected by the sex combination of parents strongly imply that Asian populations have some sort of genetic variant (or variants) which arose due to mother-child genetic conflict.

Does any of this matter? I think the article is a bit too unclear about rates of interracial marriage by group. Here is some data:

Marriage Patterns for Six Largest Asian American Ethnic Groups (Oct. 2007), Source Asian-Nation
All spouses US born/raised X US born or foreign born US born/raised X US born/raised

Asian Indians

Men (All) Male : female outmarriage ratio - 1.28 (All US born) Male : female outmarriage - 0.862
Asian Indian 91.9 73.3 56.7
Other Asian 0.9 2.7 2.8
White 5.5 18.5 31.3
Black 0.5 0.5 0.8
Hispanic/Latino 0.8 3.4 5.8
Multiracial & All Others 0.4 1.6 2.7
Asian Indian 93.6 77.5 54.2
Other Asian 0.7 1.7 2.0
White 4.3 18.9 36.3
Black 0.5 1.4 2.8
Hispanic/Latino 0.4 1.4 2.7
Multiracial & All Others 0.5 0.9 1.9


Men (All) Male : female outmarriage ratio - 0.381 (All US born) Male : female outmarriage - 0.735
Chinese 89.5 64.6 53.1
Other Asian 4.5 11.5 11.6
White 5.3 20.2 29.7
Black 0.1 0.3 0.5
Hispanic/Latino 0.7 2.1 3.0
Multiracial & All Others 0.4 1.4 2.0
Chinese 81.5 54.0 44.6
Other Asian 2.7 7.3 7.8
White 13.9 32.8 40.4
Black 0.4 0.9 1.1
Hispanic/Latino 0.9 2.9 3.5
Multiracial & All Others 0.6 2.1 2.6


Men (All) Male : female outmarriage ratio - 0.338 (All US born) Male : female outmarriage - 0.779
Filipino 82.4 50.1 35.6
Other Asian 2.8 6.9 7.3
White 9.2 27.1 36.0
Black 0.3 1.2 1.3
Hispanic/Latino 2.9 8.3 11.1
Multiracial & All Others 2.3 6.3 8.4
Filipino 61.1 37.6 28.4
Other Asian 2.8 6.4 6.9
White 27.2 40.0 46.2
Black 2.8 3.6 4.1
Hispanic/Latino 3.6 7.5 8.6
Multiracial & All Others 2.5 4.0 5.7


Men (All) Male : female outmarriage ratio - 0.516 (All US born) Male : female outmarriage - 0.829
Japanese 63.9 55.0 53.7
Other Asian 9.9 12.4 9.4
White 19.7 24.0 27.2
Blacks 0.4 0.6 0.7
Hispanic/Latino 2.8 3.6 4.1
Multiracial & All Others 3.2 4.4 4.9
Japanese 47.4 51.3 50.9
Other Asian 6.4 8.0 7.7
White 38.2 32.1 32.8
Black 1.6 0.7 0.8
Hispanic/Latino 2.8 3.0 3.1
Multiracial & All Others 3.7 4.8 4.9


Men (All) Male : female outmarriage ratio - 0.232 (All US born) Male : female outmarriage - 0.663
Korean 90.7 60.7 39.5
Other Asian 2.6 9.6 12.5
White 5.5 24.9 40.3
Black 0.3 1.3 2.1
Hispanic/Latino 0.5 1.7 2.8
Multiracial & All Others 0.4 1.7 2.7
Korean 69.4 35.3 22.5
Other Asian 3.7 9.4 8.9
White 23.7 48.9 60.8
Black 1.0 1.7 2.2
Hispanic/Latino 1.1 2.8 3.4
Multiracial & All Others 1.1 1.8 2.3


Men (All) Male : female outmarriage ratio - 0.257 (All US born) Male : female outmarriage - 0.530
Vietnamese 92.3 76.9 71.0
Other Asian 2.9 7.0 5.8
White 2.9 10.5 15.0
Black 0.2 0.9 1.3
Hispanic/Latino 1.4 3.8 5.5
Multiracial & All Others 0.3 1.0 1.4
Vietnamese 83.3 66.8 58.2
Other Asian 3.5 8.2 7.8
White 11.3 20.8 28.3
Black 0.5 1.7 2.4
Hispanic/Latino 0.7 1.3 1.7
Multiracial & All Others 0.7 1.2 1.6

More like this

23% is considered a low rate of caesarean ?

This fits with what I have read about high SES women asking for caesareans and getting them from litigation adverse doctors, worried about being sued if anything goes wrong with a natural birth or "vaginal delivery" as it is now known in some circles.

It's my impression that in a relatively high proportion of Thai/Filipino Asian woman white man couples the man is much older. This may affect things such as the incidence schizophrenia in their children for reasons that are not directly conected with geographical genetic variation.

It's my impression that in a relatively high proportion of Thai/Filipino Asian woman white man couples the man is much older. This may affect things such as the incidence schizophrenia in their children for reasons that are not directly conected with geographical genetic variation.

this is a good point. but in the USA we do have large populations, like japanese americans, where race is the primary variable you would test.

Gestational diabetes and near-diabetes makes for huge babies. You don't need a pelvic-size hypothesis to explain the increased caesarean rate - they should have looked at birthweight.

As for those observant male colleagues, it has been my observation that pelvic tilt is the real variable, not just size. In both genders. (And regional preferences for sexual position do seem to correlate - there's a reason they called it the "missionary" position).

I am skeptical about conclusion of pelvic sizes between white vs asian females. Asian, at least Chinese babies have birth weight heavier than pure Caucasian babies. So Chinese women pelvis can not be smaller than white women. Yes, white women might have more flesh outside pelvic bone which might give wrong visual impression of pelvic size. But pelive bone size is really matter for the birth. At least Rushton had some imformation about aisian babies weight.

As previous commenter, it is likely psychological factors behind the difference of c-section rates.

On the other hand, gestation diabitis is higher in Chinese women. Ironically, babie always are heavy and large in gestion diabetis. So heavy babie weight in Asians might have some thing to do with this condition.

"incidence of gestational diabetes was lowest among white couples at 1.61 percent and highest among Asian couples at 5.73 percent - and just under 4 percent for Asian-white couples."

I think we should separate actual interaction effects like hapK from cases like these where the mixed individual tends toward one parent. Case in point:

"Racially mixed individuals in Hawaii seem to be more likely to be obese;"

There were actually not many categories in that study where mixed individuals were significantly fatter than BOTH of the parent races, and you have to wonder if there is something special about Hawaiian admixture, like many did in the comments of that post. Their comparison was to the average of the two ethnic groups. For example, Asian/White, Black/White and Latino/White individuals were all closer to the heavier parent race but not significantly over in any category. I don't think that warrants the conclusion that racially mixed individuals are more likely to be obese.

I'm not going to speak to the C-Sections, that seems obvious but could the diabetes thing be as simple as Asian women married to white males tend to eat more Western food, which their bodies don't metabolize as well making them more likely to get diabetes? Asian Americans have a high obesity rate, but it is not commonly talked about.

"Asians, especially those from Far Eastern nations like China, Korea and Japan, are acutely susceptible to Type 2 diabetes, the most common form of the disease and the subject of this series. They develop it at far lower weights than people of other races, studies show; at any weight, they are 60 percent more likely to get the disease than whites.

And that peril is compounded by recent immigrants' sudden collision with American culture. Many of them left places where factory and fieldwork was strenuous, televisions were rare and advertising was limited. They may speak little English and have poor access to medical care.

Many have never even heard of diabetes, much less the recent scientific studies showing that a Western diet, high in fat and sugar, puts them in danger of getting Type 2 diabetes, which has been linked to obesity and inactivity, as well as to heredity. (Type 1, which comprises only 5 percent to 10 percent of cases, is not associated with behavior, and is believed to stem almost entirely from genetic factors.)

Many recent Chinese immigrants have come from places where food was scarce, and experts say some view fat as a trophy of wealth and status. Their children try to fit into their new country by embracing its foods and its sedentary pastimes."…

AG said:
I am skeptical about conclusion of pelvic sizes between White vs Asian females

As someone who was born in Ireland and lives in New York, and exposed to a huge selection of ethnicities daily, I would definitely say that East Asian female pelvic widths are much less than European pelvic widths in general, even allowing for overall smaller sizes of East Asians.

IMO European women display considerably different pelvic widths, based on their geographic location. In general Mediterraneans (including Jews), some Scandinavians, Northern English and South East Europeans tend to have wider pelvises, while women from the North European plain (Holland, Northern Germany, Northern Poland) and much of Eastern Europe tend to have narrower pelvises. So there are geographical differences.

. In general Mediterraneans (including Jews), some Scandinavians, Northern English and South East Europeans tend to have wider pelvises, while women from the North European plain (Holland, Northern Germany, Northern Poland) and much of Eastern Europe tend to have narrower pelvises. So there are geographical differences.

LOL. wow, you're specific paul.

could the diabetes thing be as simple as Asian women married to white males tend to eat more Western food, which their bodies don't metabolize as well making them more likely to get diabetes?

No, the risk is lower for Asian women having babies with white men than for Asian women having babies with Asians (4% vs 5.73%)

I'm filled with admiration for Paul's powers of observation, but it's a bit of a worry that he's been taking such detailed notes :)

What really matters is the size of the baby's head in relation to the size of the birth canal. Pelvic size, or at least hip width, may not be a perfect proxy for the size and shape of the birth canal, and birth weight may not be a perfect proxy for head size. Confounding factors could be the incidence of breach births, and the fact that some women elect to have caesarian deliveries for cosmetic reasons or to avoid the intense pain of vaginal delivery. Plus Tod's observation about high SES women.

African women have narrower hips than European women, but their pelvises are deeper. Babies' heads are plastic to some extent. And the pelvis does expand somewhat during birth, and never completely resumes its original size.

Obesity and lack of physical fitness in the mother are other possible confounding factors. Slim, fit women have less difficult births. Asian women may be slimmer and smaller, but they might not be fitter physically.

It also depends how they have grouped people racially. I don't have Paul's notes in front of me, but in my extensive personal observation, Indian women tend to be generously hipped, maybe even more so than those big-arsed northern English sheilas. For the purpose of examining mother's pelvis size as a factor in birth difficulties in racially mixed marriages, grouping them with Asians would be confounding, and for that purpose at least I would be classifying a European male/Indian female marriage and vice versa as "not mixed" and grouping Indians and Europeans together.

I'm not saying it's wrong necessarily, but it is very broad brush.

By Sandgroper (not verified) on 03 Oct 2008 #permalink


Slightly off topic, as I see your point, but your example of African Americans having heightened risk of heart problems is still up for debate. As was discussed last year on the old site, it appears hypertension is only higher in African Americans, but not in other admixed black New World Populations, like Jamaicans, and African Americans are still lower than many Western European groups. It's obviously something more complex...some thought maybe it has to do with latitude...…

Vietnamese are a small people with big heads for their body size.
It is my impression that Koreans have very large heads. There are SES differences too.
A study said to refute suggestions of the adverse effects of interracial mating was criticised for using Hawaii on the grounds that the average "black" person there has a lot of non-black ancestry.

Perhaps this type of analysis could be the next significant revenue stream for companies like Decode and 23andMe - a dating site where you are pre-matched to potential partners, chosen with the intent of maximizing offspring viability.

Is this something that can be done today, or is the technology not 'there' yet?

By Stagyar zil Doggo (not verified) on 05 Oct 2008 #permalink

"Although the probability of heart attack is raised in people who are overweight, have a thick waist relative to their hips, smoke and are in a low socio-economic group, statistical tests suggest that finger ratios predict age at heart attack independently of these life-style factors. However these same tests suggest that lifestyle factors remain independent indicators of raised risk of heart disease. Male mean ratio for a sample of Jamaican men is 0.93 "Female type finger ratios are related to high levels of fibrinogen and thyroid stimulating hormone"
from The Finger Book (2008 Faber and Faber)by John Manning

check out genepartner,

they don't, as of yet anyways, advertise checking for pelvic to infant head ratios. but maybe u could convince them otherwise.

Sandgroper, Wongba,

I take pelvic size seriously, I have to!

My Mom's family hails from Northern England, and most of the women have hour-glass figures, with extra wide pelvises. The family also has a tendency to have large heads - but with this combination, no problems.

My daughter's French/Italian mother was on the more athletic side - her dad was an olympic sprinter for France - and had a somewhat narrower then average sized pelvis.

At birth our daughter's head was in the 97.5% for size, and was a full 2 inches wider then her mother's pelvis - so she had to be delivered via C-section. In earlier times, one or both may have died in childbirth!