A Blog by Ed Yong

Genes and culture: OXTR gene influences social behaviour differently in Americans and Koreans


There are great plays and bad ones, but the playwright’s actual text is only one aspect of a production. The very same words can take on radically different meanings depending on the whims of the director, the abilities of the actors and the setting of the stage. The same is true of our genes and our environments. In cases where genes affect our behaviour, the same stretch of DNA can lead to very different deeds, depending on individual circumstances. Just as a production defines a play, environments and cultures alter the effects of certain genes.

Heejung Kim from the University of California has discovered a great example of this effect by studying a gene called OXTR (or the ‘oxytocin receptor’, in full). The gene creates a docking station for a hormone called oxytocin, which is involved in all sorts of emotions and social behaviours, from trust to sexual arousal to empathy.

Kim looked at a specific version of the OXTR gene, whose carriers are allegedly more social and sensitive. But this link between gene and behaviour depends on culture; it exists among American people, who tend to look for support in troubled times, but not in Korean cultures, where such support is less socially acceptable. Culture sets the stage on which the OXTR gene expresses itself.

OXTR varies from person to person, and the DNA ‘letters’ at particular spots can affect the way we behave. According to previous studies, people with a ‘G’ at one specific site tend to be more sensitive parents, more empathetic and less lonely than those with an ‘A’. But most of these studies have been done with white, Western people who are hardly representative of the world at large – in fact, they’re positively W.E.I.R.D.

To looked outside this “thin and rather unusual slice of humanity”, Kim compared 134 Korean students with 140 American ones, all with comparable splits of age, gender and background. Using a questionnaire, she measured how stressed each volunteer was feeling at that point in their lives, and how they cope with stress. As with previous studies, Kim found that Koreans are less likely than Americans to turn to their social circle for support and they get less out of doing so; they are more concerned about burdening their friends and straining their relationships.

The OXTR gene exerts its influence against the background of these contrasting cultural conventions. Distressed Americans with one or more copies of the G version were more likely to seek emotional support from their friends, compared to those with two copies of the A version. But for the Koreans, the opposite was true – G carriers were less likely to look for support among their peers in times of need (although this particular trend was not statistically significant). In both cases, the G carriers were more sensitive to the social conventions of their own cultures. But the differences between these conventions led to different behaviour.

And in a further example of the influence of the environment, Kim only found this pattern among people who were experiencing a lot of stress. In the low stress group, she found that Americans were indeed more likely to seek emotional support than Koreans, but their OXTR gene had no bearing on their choices.

Of course, Koreans and Americans differ not just in their cultures, but in their genes (including many others beyond OXTR). To account for that, Kim also worked with a small group of 32 Korean-Americans who were born and raised in the US, but were genetically Korean. Kim found that the link between OXTR and emotional support among these volunteers was much closer to the culturally similar Americans than the genetically similar Koreans.

Richard Ebstein, who has worked on OXTR before, says, “Overall, I would say it’s a very interesting finding…  These types of studies are needed to help us get a better understanding of how it’s not just nature or nurture but rather the interplay between the two that contributes to how we deal with the social environment.” However, he’s not convinced (and nor am I) that Kim looked at enough people, particularly in the extra experiment with the Korean-Americans. Ebstein wants to see them repeat the results in a much larger group.

Even so, Kim’s results are compelling. They’re also unusual in looking for an interaction between genes and culture. Many studies have looked at how nature and nurture work together but in most cases, the “nurture” bit involves something social that’s either harsh or kind, such as loving or abusive parenting. In one of the most famous examples, people with the ‘low-activity’ version of the MAOA gene tend to be more aggressive than those with the ‘high-activity’ one, but only if they’ve been abused or neglected as children. Kim’s study stands out because it looks as cultural conventions instead, and Ebstein says that it “provides an interesting new avenue for researching gene-environment interactions.”

Kim also hopes that her work will encourage more scientists to investigate the ways in which genes and culture evolve together. She notes that the G version of OXTR is more common among white Americans than Korea. It’s tantalisingly possible that American culture has come to emphasise social support partly because more people have genes that skew them towards social behaviour. So genes constrain culture, while culture creates the stage on which genes exact their influence.

Reference: PNAS http://dx.doi.org/10.1073/pnas.1010830107

More on genes and environment:

If the citation link isn’t working, read why here

Twitter.jpg Facebook.jpg Feed.jpg Book.jpg

13 thoughts on “Genes and culture: OXTR gene influences social behaviour differently in Americans and Koreans

  1. ” Kim found that the link between OXTR and emotional support among these volunteers was much closer to the culturally similar Americans than the genetically similar Koreans.”

    Did the Korean-Americans have the same average OXTR profile as the Korean-Koreans? I wonder to what extent immigration selection is involved in findings like the above. I have known a few Sea Turtles that have returned, because they didn’t feel they matched the life style in the US.

  2. Somehow I strongly doubt that the link between genetics, especially a single gene, and behaviour is as close as this research wold indicate.

  3. And how close is that? Note that every time I talk about an influence of the OXTR gene on behaviour, I always say “more likely to” or “less likely to”. It’s not the case that the G version leads to social sensitivity and the A version doesn’t. Nor is it the case that OXTR is the only gene involved in social sensitivity – far from it.

  4. Very interesting finding, would like to read more about it. Unfortunatelly I can’t link to the PNAS reference and I’ve had a tough time finding even an abstract on the web. Any help would be much appreciated!

  5. I am thinking that genetic influences and cultural influences are independent of each other and both present in our decision making process.
    Sometimes these influences oppose.
    So I understand genetic influence cannot change its character.

  6. Ed Yong said, “In one of the most famous examples, people with the ‘low-activity’ version of the MAOA gene tend to be more aggressive than those with the ‘high-activity’ one, but only if they’ve been abused or neglected as children.”

    Wrong! You should really get a better handle on the research before making a blanket statement like that. I tried to correct you before, but you made the same stupid mistake.

    There is no such thing as a “‘low activity’ version of the MAOA gene.” Most of the research on MAOA compares the 3-repeat allele of the VNTR promoter to the 4-repeat allele because those alleles are conveniently common. The 2-repeat allele is lumped in with the 3-repeat allele, which defines the “low-activity version,” but it is so uncommon in white people that it could be excluded without impacting results. “MAOA-L” is purely a mental construct.

    The 2-repeat allele seems to be about an order of magnitude more common in black people. It is, if you will, a lower-activity allele, which is associated with a doubling of violent delinquency compared to the 3- and 4-repeat alleles IN THE ABSENCE of any gene-environment interaction (Guo et al, 2008).

    Brunner syndrome is an even lower-activity allele caused by a mutation that completely inactivates the gene. The syndrome is associated with conduct disorder, rape, and arson. Fourteen people have been found to have the syndrome. Only four people have ever been found with the mutation but lacking the syndrome.

    The only reason that the 3-repeat allele requires an environmental interaction to be associated with violence is because it is, in effect, only mildly different from the 4-repeat allele. In fact, other gene-interaction associations with violence work just as well for this allele. For instance, it interacts with the hormones, testosterone, estradiol, and cortisol. Plus, Beaver et al just created a crude “genetic index” that included MAOA, along with four other genes. This index was associated with black violence EVEN IN THE ABSENCE of a gene-environment interaction.

    This is important research that is now starting to influence the criminal justice system of the United States. Many falsehoods have been written about this gene in order to defend political correctness or to pretentiously dismiss genetic determinism. You are one of many guilty of this, and you really should stop now.

  7. Far too much judgment based on far too little data, interpreted in accordance with whatever the researcher had on his mind at the time.

    And in any case, correlation is not a demonstration of causation. To put it simply, the results of this study are worthless, and the effort put into it wasted. The researcher may be right (though typically there are many more ways to be wrong than to be right, so wrong is the more common outcome) or he may be wrong, but his study and results tell us nothing for or against his conclusions.

Leave a Reply

Your email address will not be published. Required fields are marked *