Why the Chinese don’t buy deodorant

In human populations a SNP in ABCC11 is correlated with two salient traits: 1) wet or dry earwax 2) body odor. When I had my first son sequenced before his birth the main variant of phenotypic consequence that I noticed (aside from him being a heterozygote on KITLG), was that he carried a derived mutation on this position. Meaning that he was going to have dry earwax and fewer issues with body odor.

My wife and I are both heterozygotes. This is not too surprising. The derived variant is actually greater than 50% in Bengalis in the 1000 Genomes (in South India the derived variant is also around ~50%), while about ~25% of Northern Europeans are heterozygotes.

This genetic story came to my mind again because of this article in The New York Times, Aiming at China’s Armpits: When Foreign Brands Misfire:

There’s another reason few Chinese consumers buy deodorant: basic biology.

Scientists in recent years have shown that many East Asians, a group that includes China’s ethnic Han majority, have a gene that lowers the likelihood of a strong “human axillary odor” — scientist-speak for body stink.

That lowers the likelihood that they will use deodorant to begin with, according to a 2013 study by researchers at the University of Bristol and Brunel University in Britain, after a survey of nearly 6,500 women of various backgrounds.

“It is likely that deodorant usage is not widely adopted because there is, for much of the East Asia population, no need for it,” it said. (For those curious about such matters, that same genetic difference also leads to drier earwax.)

A friend of mine in undergrad of East Asian background told me once that she had never worn deodorant. So this shouldn’t be very surprising.

Today I found a paper, A missense variant of the ABCC11 gene is associated with Axillary Osmidrosis susceptibility and clinical phenotypes in the Chinese Han Population, which explicitly probes the correlation between body odor (“Axillary Osmidrosis”) and the SNP in question in the Han Chinese population.

The chart below makes the association obvious:

The correlation between carrying the G, ancestral, allele, and body odor is very strong. Though it is imperfect. Going through this literature human smells are clearly a polygenic trait (see The effect of ethnicity on human axillary odorant production). That being said, this case-control study in a Han population shows ABCC11‘s importance in at least East Asian populations (earlier work in Japan showed that those with body odor tended to have wet earwax and carry the G allele as well).

In regards to the genotype proportions the authors observe:

The excessive heterozygosity observed in AO individuals is probably due to the effect of selection, particularly nonrandom mating against AO phenotype.

This doesn’t make sense to me. Wouldn’t people who have body odor tend to pair up in a society where they are a minority? The authors note that the excess of heterozygotes was observed in earlier studies too.

If you dig into the frequencies it seems that the derived mutation is absent among populations in Africa without recent Eurasian back-migration. I looked it up, and it’s segregating in ancient Eurasian samples, with Ust Ishim being a heterozygote. It is curious that in no population has the derived frequency swept to fixation, nor has the ancestral variant fixed in other groups (such as in Europe).

I strongly doubt that there is any selection on this locus due to earwax or body odor. It is a pleiotropic locus, there are other effects from the mutation. One of those other effects is probably the target of any selection. And in regards to selection, it seems likely that that would be a balancing sort since neither the ancestral nor the derived variant are fixed in most populations.


18 thoughts on “Why the Chinese don’t buy deodorant

  1. It amuses me that they invented ethnicity to avoid saying “race”, and now they just use it as a euphemism for “race”. The effect of ethnicity on body odor would be from things like diet, hygeine customs, and religious practices.

    One of my friends said to another “hey, you’re ethnically Italian, aren’t you?” Maybe she should have said “no, I’m ethnically American, if you want to say I’m racially Italian, say so.” But she knew what they meant really, so she just nodded confirmation.

  2. It’s indeed ironic that people who, understandably, want to keep the concepts of race and culture separate suddenly start conflating them again by using the word “ethnicity” to mean race.

    When I lived in Japan I had to beg my mother to send me toiletries care packages–not only deodorant but shampoo and conditioner, since their products turned my hair into a stringy, greasy mess.

  3. Are those American populations in the figure natives or mestizos? If natives, the distribution of the alleles would appear to indicate that 1) the Asian variant appeared before 14,000 cal BP and 2) as other genes also show, ancestral Beringians/Paleoindians were ca. 40% ANE and have been replaced by Han-related populations in NE Asia. Curious, though, that EDAR distribution is very different–fixed in Native Americans (along with shovel-shaped incisors).

  4. I do wonder if there is some effect such for ABCC11 the derived variant gains the selective advantage, once there is some other change upstream in East Asian genetics, one which may have been a mutation which occurred more recently and/or is more restricted to East Asia. For instance, some effect on gland development from the derived EDAR variant then makes ancestral allele carriers at ABCC11 who have the derived EDAR variant less fit.

    With selection on ABCC11, also skeptical on the odor element itself being selected. As well you could see a selective story on negative body odor, but if any of the studies showing positive romantic attraction benefits from the “right” body odor check out (while it may be none of them do!), then that should disadvantage derived variant carriers at least a bit as well, and the balance of these forces is more questionable.

    @Stuart, per the post, Ust-Ishim is heterozygous with a derived and ancestral allele, so the Asian modal derived variant had to have appeared and be at reasonable frequencies in some early Eurasian populations by at least 45,000 kya (long before the genetic founding of the Native American population, and IMO likely before various Eurasian streams had coalesced into the population that would become East Asians).

  5. Matt,

    The relationship, if any, of Ust-Ishim to later Asians seems to be an unresolved issue. It is frustrating that few relevant ancient genome studies (e.g., of the Anzick infant or Shuka Kaa) provide information about EDAR variants in early Native Americans. Based on back-projection from modern samples its origin in East Asia is dated to about 25,000 BP (see most recently Smith et al. in BioRxiv).

    Could body odor have any selective effect relative to parasites or mosquitoes?

  6. The wet earwax allele is statistically associated with breast cancer risk in Japanese women, but apparently not in European women.

    Recent research shows that malaria-infected mosquitoes are drawn to human body odor. The same chemicals are used by some orchids to attract mosquitoes for pollination.

    I remember reading an article, many years ago, in which the Leakeys speculated that the big predators around Olduvai Gorge left them alone because they were repelled by their scent.

  7. I remember first hearing about this from an east Asian American school mate in law school. “Asians don’t stink,” he said. I was skeptical at the time because I had a class sitting next to a Chinese student from some back water city who had very poor hygienic habits who not only stank, but had terrible breath (atypical of the Chinese LLMs, I should add).

  8. Many interesting items here, at Alex Kim’s blog


    including this:

    Jun Ohashi, Izumi Naka, & Naoyuki Tsuchiya (2011)

    Molecular Biology and Evolution 28(1): 849-857; doi: 10.1093/molbev/msq264

    A nonsynonymous single nucleotide polymorphism (SNP), rs17822931-G/A (538G>A; Gly180Arg), in the ABCC11 gene determines human earwax type (i.e., wet or dry) and is one of most differentiated nonsynonymous SNPs between East Asian and African populations. A recent genome-wide scan for positive selection revealed that a genomic region spanning ABCC11, LONP2, and SIAH1 genes has been subjected to a selective sweep in East Asians. Considering the potential functional significance as well as the population differentiation of SNPs located in that region, rs17822931 is the most plausible candidate polymorphism to have undergone geographically restricted positive selection. In this study, we estimated the selection intensity or selection coefficient of rs17822931-A in East Asians by analyzing two microsatellite loci flanking rs17822931 in the African (HapMap-YRI) and East Asian (HapMap-JPT and HapMap-CHB) populations. Assuming a recessive selection model, a coalescent-based simulation approach suggested that the selection coefficient of rs17822931-A had been approximately 0.01 in the East Asian population, and a simulation experiment using a pseudo-sampling variable revealed that the mutation of rs17822931-A occurred 2006 generations (95% credible interval, 1,023–3,901 generations) ago. In addition, we show that absolute latitude is significantly associated with the allele frequency of rs17822931-A in Asian, Native American, and European populations, implying that the selective advantage of rs17822931-A is related to an adaptation to a cold climate. Our results provide a striking example of how local adaptation has played a significant role in the diversification of human traits.

  9. “Wouldn’t people who have body odor tend to pair up in a society where they are a minority?”
    Probably depends also on the hygienic conditions. If everyone stinks badly, the differences might not be great enough to be noticed. Personal hygiene is still not the strong side of Chinese (men in particular) & it was probably much worse in history, with the exception of higher echelons.

  10. “Personal hygiene is still not the strong side of Chinese”

    You research on that must have been truly staggering in scale and geographic extent, but somehow managed to miss most of the Chinese I have ever known (really a lot).

  11. Interestingly enough (for me anyway), dogs that belong to the Northern Spitz-type, e.g. Siberian Huskies, Samoyeds, Akitas, Jindos, etc. have very low (almost imperceptible to humans) body odor, and are also extremely averse to soiling their habitat. These are, of course, all cold climate dogs. Given that dogs and people have co-evolved, I wonder whether there is a connection, a commonly shared selection mechanism.

  12. “Personal hygiene is still not the strong side of Chinese”

    I can’t say I disagree in the absolute sense, but I would also speculate that their personal hygiene per affluence level is probably much higher.

    Japanese and Koreans used to be quite unhygienic back in the day, but are probably obsessively clean by Western standards today. Both societies consume massive amounts of disposables, because they loathe to touch things others touched. Even housing!

  13. On the tangent of the history of grooming in East Asia, I think I can remember hearing that generally the history of perfume and cologne, etc. was a little truncated in East Asia. Missionaries and traders in Japan reporting not much use of scents by native Japan during early encounters. Much in contrast to European encounters in the Middle East and South Asia, where these were huge trade goods.

    That’s probably got a similar explanation to lack of demand for deodorant.

    But a slight twist in the tale is there was a nice little study in 2012 that found that people tend to select perfumes which interact with their body odor to enhance attractiveness (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3314678/).

    If people tend to lack body odor in NE Asia, they couldn’t really do this in the same way; perfumes would smell nice but not work with their own odour to enhance it. Hence another explanation of lack of demand – not just that there’s nothing to mask, but also nothing to enhance.

  14. The results you cite are familiar. The open question remains:

    What has led to selection/fixation of Asian EDAR and ABCC11 variants?

    Or, is this simply a Founder effect boosted by mating selection on variants that are pretty visible once they reach a tipping point?

    My guess would be the former, since that is what one usually discovers when you look into these things, but I couldn’t tell you precisely why.

    It is a particularly vexing challenge because it can’t be selecting on something that is universally fitness enhancing. It has to be something that providing an extreme selective fitness advantage at a particular time and place for a particular founding population or some particular founding populations (whose later success may be due to cultural or technological advantages rather than genetically driven phenotype advantages). But, that was not strongly selected for at almost all other times and places.

    But, explaining EDAR is very hard with any plausible “just so story” (one can imagine several for ABCC11 more easily, particularly if you assume a Northern or clockwise migration route from Africa to Asia). And, even the most popular “just so stories” for selection on LP in Europe, which are straightforward to propose, have trouble explaining why selection was so intense and so rapid on this gene.

  15. IF the derived ABCC11 allele is cold-adapted and appeared ca. 50-60,000 years ago, and IF there was a millennia-long “Beringian standstill”, one would think the derived allele would have become fixed in Native American ancestors. Obviously that did not happen, so something else is going on.

Comments are closed.