When you choose to publish with PLOS, your research makes an impact. Make your work accessible to all, without restrictions, and accelerate scientific discovery with options like preprints and published peer review that make your work more Open.


20 Gene Variants and Transgender Identity: What Does It Mean?

The week started strangely.

On Monday morning, the author of a new book on transgender identity emailed me, asking about my research (I don’t have any). She’d read my comments in The Daily Mail about an abstract from a meeting identifying gene variants associated with transgender identity in a handful of people. But the writer, with whom I hadn’t communicated, didn’t name the researcher. So people thought it was me.

Soon, Google Alerts sent me the full version of the article in The Times, by Oliver Moody, and then I recalled having emailed with him. (This is the Times of London; the New York Times rejects anything I send.)

Since then I’ve been quoted in several outlets with whom I never corresponded – such is the nature of echo-chamber journalism these days – and I’ve had requests from people seeking genetic testing for transgender identity. It’s way too soon!

How Hype is Born

I deduced that Mr. Moody had contacted me after reading “Is Transgender Identity Inherited?” here at DNA Science from about a year ago, or had heard an NPR interview follow-up. The end of my post states that “a genomewide SNP screen, or even genome sequencing, might one day identify a genetic ‘signature’ for individuals more likely to identify as transgender.”

In the new abstract, “Understanding the Genetic Basis of Transgender Identity,” presented at the Society for Reproductive Investigation’s annual meeting in San Diego March 6-10, J. Graham Theisen, MD, from the Medical College of Georgia at Augusta University and colleagues described a strategy midway between my suggestions. They sequenced exomes (protein-encoding portions of the genome) of 30 individuals who identify as transgender.

When Mr. Moody sent me the abstract last month, I was ready to tear it apart – but I was impressed. Hence my quote, “These are highly reputable folks going about this exactly the right way, searching the genomes of transgender people to highlight which genes they have variants in. It lends legitimacy, if that needs to be added, that transgender is not a choice but a way of being.”

Alas, despite Mr. Moody’s emphasis on the preliminary nature of the findings, and a spot-on lede (“Transgender men and women may carry genetic variants that influence their gender identity, a study suggests,”) his article fell victim to hyped editorial packaging.

The headline “Science pinpoints DNA behind gender identity” is absolutely untrue. Worse is the image beneath the headline of a stereotypical bespectacled, lab-coated old white dude pointing seriously to a DNA sequencing gel, with the caption “The DNA of those with gender dysphoria was found to have nine variants that may cause the condition.” Cause? I don’t think so. And the project leader, Dr. Theisen, actually looks like a young Chris Reeve (aka Superman), not the older guy.

Looking for Genetic Clues in 30 Individuals

The researchers sequenced the exomes of 14 female-to-male and 16 male-to-female people identifying as transgender. In the “round-up-the-usual-suspects” approach used for various types of genome-wide studies, the investigators began with variants of a logical group of candidate genes (which regulate steroid metabolism in the brain during development) and are rare (fewer than 1 in 10,000 individuals) in the general population. A database called ExAC compiles gene variant frequencies from many research groups.

Would the selected rare gene variants show up among the participants, and if so, would that trend persist in a much larger group? Then those variants might be associated with some aspects of transgender identity. But it’s a long way from association to correlation to causation. Some comments from readers of the Times piece pointed out the small sample, but conducting a pilot study is how science works. Funding agencies aren’t about to fork over the pounds or dollars for a large-scale investigation without a preliminary look-see.

The team whittled down 120,000 genetic variants of possible interest among the participants, then selected 30 variants, representing 20 genes, for further investigation (the exact figures haven’t been analyzed yet). Eleven of the variants changed one amino acid in the encoded protein to another (a missense mutation), and 9 variants were deemed “pathogenic” for neurological development and/or sex steroid metabolism – adding up to the 20 of interest.

“Pathogenic” is genetics jargon for a gene variant that has an effect—it doesn’t mean that transgender identity is a disease. I have a few pathogenic gene variants myself – we all do. Most are recessive and present in only one copy (heterozygous) so are never noticed. The researchers are still re-sequencing the candidate genes for confirmation, but so far only 2 variants are present in two copies (homozygous) in anyone.


The conclusion is clear: “We identified genetic variants in 20 genes that may play a role in transgender identity.” But it is still a powerful statement, for it lends justification to the idea that identifying as transgender is neither a mental illness nor a lifestyle choice – it is a variation on the theme of being human, as we all are in one way or another.

Dr. Theisen cautions, “Our research is still preliminary in nature, and we are a very long way from knowing what (if any) significance our findings have. The DNA of any randomly chosen individual will contain unique variants, and as a result, identifying variants alone is not sufficient to make any claims regarding causation of, or even relation to, a given phenotype. Therefore, we have not yet identified any particular genes or genetic variants that we know to be involved in the development of transgender identity, though that is the eventual goal of our research.”

The Bigger Picture

I’m not an expert on gender issues by any means, despite my being quoted this week in Gay Times and LGBTQ Nation. But I think and write a lot about genetic information. Would genetic testing for transgender identity do harm or good?

Consider the categories of DNA tests: ancestry, forensics, health, and traits. Ancestry and forensics aren’t relevant, so that leaves health-related genetic tests and traits.

If Dr. Theisen’s team or another validates a panel of gene variants associated with transgender identity in a larger population, a few things could happen.

Companies will spring up, like mushrooms after a spring rain, to package and pitch products to consumers. This could happen with more thought to economics than to the potential problems that might arise, such as a parent sending off a spit sample from a young child convinced she is a he or vice versa.

Then, with more research, if the association becomes a correlation or even a cause (such as zeroing in on a particular mutation), might health insurers cover a greater variety of transitioning-related procedures? Again, it’s economics versus consequences. Could expanded insurance coverage medicalize transgender identity, and would that counter the change of “gender identity disorder” to “gender dysphoria” in the 2013 Diagnostic and Statistical Manual of Mental Disorders (DSM-5)? The new designation denotes individuals who are distressed by the mismatch between their gender identity and anatomy, but it keeps others who identify as transgender out from under the psychiatric umbrella.

A better alternative would be for DNA tests for transgender identity to fall into the category of what direct-to-consumer genetic testing company 23andMe lists as, simply, traits. Their offerings are both popular and ridiculously obvious, like eye color, freckles, dimples, cleft chin, unibrow, and hair color, loss, and texture. Findings support what a person already knows from looking in the mirror.

Isn’t the same true of gender identity? Won’t a profile of gene variants only confirm what a person already knows?

I eagerly await publication of the full paper on the work from Dr. Theisen and colleagues.









  1. Surely one of the commercial responses to this is likely to be the sale of kits that will test whether a foetus has these markers, with the aim of enabling parents to avoid having trans children?

  2. I’m afraid you are right, Natacha. I’d hope there would be enough backlash to shut such a company down. Prenatal testing like this would further blur the lines between genetics and eugenics. Thanks for your comment! My mind didn’t even go there, but I can see this as a possibility.

  3. It would be nice to see conservatives suddenly supporting abortion, honestly. Being a trans kid who commits suicide due to their crappy conservative parents is worse than never being born.

    1. Wow. What a horrific comment, Clio. Very Nazi Germany of you. They might have a bad life so let’s just kill them? Gosh. I pray you are no one’s parents. As a scientist this is piss poor. Also, if we do find a genetic variation responsible for all or the bulk of real transgender, then that would be a candidate for genetic correction with CRISPR, not someone to murder before they are born. Sheesh.

  4. Your comment is ridiculous. Some people have abortions because they do not want a female or Down syndrome child. It is not because they believe in having only white, Christian babies. If you are okay with killing undesirables with Down syndrome because their lives might be more difficult, then you should be for culling a transgender child because they have higher suicide rates. As a scientist I am tired of other scientists and their myopic view of the world. Just bash conservatives because I live in an echo chamber. Whatever happened to doing the hard work of understanding more than one perspective?

  5. To which comment specifically are you referring? Clio’s or mine? Your post says in reply to mine. I appreciate your taking the time to comment, but a blog post is not meant to necessarily share all perspectives, and anyone is entitled to an opinion. Were you implying that I do not work hard, or one of the people commenting?

  6. Exactly.

    When it comes to abortion, who’s morality is superior?

    That of a religious Muslim or Christian person wanting to abort a fetus containing the genetic markers for trans or gay, if such a test was hypothetically available?

    A Chinese or Indian woman aborting a female because she and her family culturally value boys over girls?

    Or a Western woman wanting to abort a fetus with Downs markers, or simply because she’s starting to grad school next semester?

  7. This comment: “at least here in the US where some folks think “human” means a white, male, Christian heterosexual.”

    What purpose does this comment serve?

    My opinion is that too many people will make comments like that because it is the easier thing to do. I just don’t think bias is the best attribute of a scientist. Your comment lacks any sort of affirmation of the intricacies of eugenics and the fact that it is not just a “white, male, Christian heterosexual” who might conceive of aborting a transgender fetus. You can have an opinion, but I expect better of someone with a PhD than to just say what they think their audience would like to hear. I have no clue how hard you work, but merely have the opinion that such a snide comment might give someone the idea that you have a very lazy, narrow perspective. It certainly would affirm to a “white, male, Christian heterosexual” that scientists are the enemy and that defunding NIH is a good idea.

  8. In the West aborting fetuses – usually female – based on sex is illegal, so there is a general moral consensus as well as legal codification based on that ethos. The individual choice of a woman wanting to do is overridden by law and condemned even by progressives, who apparently consider their morality in that case to be superior.

    Abortion is an ethical thicket for anyone who looks at it honestly. It’s not just about simple minded, medeival Christians being against “choice” and progress.

  9. I apologize Matt, I misunderstood, I thought you were criticizing my blog post. You are referring to an off-the-cuff comment about someone else’s comment. Is there something you object to in the post itself? It isn’t easy to accommodate all perspectives in an informal, 1000-word post. I cover eugenics thoroughly in my textbooks. At times, in the US, older white males do seem to be running things. That’s all I meant. No need to belittle me or what I do.

  10. Conservatives and other humane human beings do not support abortion because it is immoral and unethical to kill any child – born or unborn.

    “Being a trans kid who commits suicide due to their crappy conservative parents is worse than never being born.”

    Who are you to decide for a transgender child or anyone else what is best for them? If your perverse theory is correct, then at least the child would have the option of suicide. What about the other children who prefer life over death? Do you think we should have them all put down? A child who is born at least has the chance to decide for him or herself.

  11. Thanks for sharing your opinion Dinah. This is Ricki. I wrote the blog post. I DID NOT write the comment to which you refer. That was a comment, and it appears to be gone although I did not delete it. (The site can get wonky). I don’t see it but I remember reading it, and I think the person was writing about their own experience, not necessarily dictating what others should do. I repeat, I did not write that. Your comment addressed the blog in general.

  12. Thank you for this very interesting article. I really appreciate your helpful analysis of both the science and the social issues surrounding its communication.

    I feel there are some additional potential issues with the research should “transgender” gene variants be identified, in addition to those you outline.

    There is a risk that if some trans people are easily identified through these gene variants and others will not, a hierarchy of “realness” may be created – one which doesn’t necessarily map on to people’s lived experiences and needs, but is nevertheless used to determine both social recognition and access to healthcare.

    Additionally, there will be fears within trans communities that people will seek to identify whether or not their child is trans prior to birth, with unwanted trans pregnancies terminated accordingly. Plus, if trans people become more easily identifiable through testing, this may put people at great risk within some anti-trans authoritarian states.

    On a slightly different note – in my own (social) research, one of the overwhelming messages from trans patients is that healhcare providers of all professions (including gender identity specialists!) are often not at all interested in listening to “what a person already knows” about their lives and experiences. I’m not sure making trans people more easily identifiable would actually usher in the social changes necessary for trans people to recieve more respect and be listened to within healthcare settings.

  13. Thanks so much Ruth. The part you mention that healthcare providers ignore — the certainly of a person’s story — is what I find the most fascinating, and also compelling if non-technical evidence that transgender identity is innate, part of who someone is, and a variation, not a pathology. I agree that any sort of genetic testing has the potential for inflicting harm.

  14. There’s also a whole class of bigots out there who denies that transgender women are women, and they think that transgender men are sex traitors transitioning to enjoy the privileges that they feel comes with being male. They think that biological sex is genitals and chromosomes only. That Gender doesn’t exist, that it’s made up by society. Transgender Men and Women directly contradict their mistaken beliefs, so they attack us with every fiber of their being.A scientifically proven biological or genetic factor could serve to end much of that. It can’t come quick enough.

  15. Thanks for writing Susan. I was on the receiving end of such a bigoted attack, as you can see in these comments. Trolled for a comment I made about a comment, not even the article. Are you THE Susan of susans.org? I just discovered the website last night. I will be writing about this again so if you hear of any new research about anything biological, please get in touch (www.rickilewis.com). Thanks again.

  16. Really interesting article, are you looking at widening your sample set which was quite narrow. If so I would be willing to share my DNA profile from 23 and me.

    I agree with Ruth Pearce above, sometimes it is just good to know if your genotype aligns.

    It would also be interesting to look at some of the other alignment to other research on phenotypical responses to genotypes, to see if there is a true correlation, like the one on “foetal testosterone wash” influenced by genetic sequences and if it displays as a phenotypical response characterised by the length of the ring and index finger.

  17. Dr. Theisen is enlarging the sample size. His reported work was just an abstract from a meeting, and a journalist learned of it and publicized it, perhaps too soon. A pilot study is pretty typical. Yes, the digit ratio phenotype is fascinating! Thanks for posting.

  18. It’s funny how no one has brought up that there is a great possibility that people will use the test to prove either they or their child are delusional, wishful thinkers and not transgender. I for instance have a daughter who acted just like I did growing up, but thanks I believe to the culture she was raised in for a time, suddenly decided she was a boy. Just before she had said she was a lesbian. She has always looked for attention. . .so this test would settle a long argument that has nearly destroyed our relationship. In fact, she has nearly disowned all her family over her beliefs!! We love her, but we don’t believe her ideology. She also is an Atheist, but she can handle disagreements there apparently.

Leave a Reply

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

Add your ORCID here. (e.g. 0000-0002-7299-680X)

Back to top