The data on twins suggests that there is a genetic component to gender dysphoria.
It also suggests that other factors are involved in developing gender dysphoria.
Unfortunately, the data is weak because it is mostly made up of case studies.
In addition, the data on identical twins and the data on fraternal twins were collected in different ways.
It is possible that this review overestimates the influence of genes due to the way the data was collected.
I had thought that writing this review would be quick and easy; genes are involved, but they are not the only factor. The truth is that the data is flawed and we don’t have conclusive proof yet. What we do have is a suggestion that genes are involved in gender dysphoria and a need for more research in this area.
Data on twins with gender dysphoria is hard to collect because it is rare. This review is an important one and it shows that there are good reasons to keep looking for possible genetic links to gender dysphoria. It also shows that there are good reasons to look for non-genetic factors that play a role in developing gender dysphoria.
Now you have the summary of the results, back to the study. Why does it mattter if many of the reports on identical twins came from case studies?
Using case studies means that there is a possibility of selection bias.
People may be more likely to publish interesting cases. For example, the review includes one case study where both identical twins had gender dysphoria, but only one had schizophrenia. In another case study both identical twins had anorexia, but only one had gender dysphoria. These cases are interesting, but they may not be typical.
This can become a more serious problem if therapists are more likely to be interested in cases of twins who are both trans. Alternatively, there could be a selection bias in favor of writing about identical twins where only one twin is trans. Some therapists might unconsciously look for cases of twins that fit their own theory about the cause of gender dysphoria. (Read more about case studies and selection bias here.)
The results of this review suggest that there is a selection bias that favors identical twins over fraternal twins. In other words, people write up and publish cases of identical twins more often than fraternal twins.
This is not because trans people don’t have fraternal twins; the studies that reviewed clinic records found 19 sets of fraternal same-sex twins and only 7 sets of identical twins. Only 27% of the twins in this group were identical twins. This is not surprising; fraternal twins are more common than identical twins in the general public.
The other studies, in contrast, reported on 16 sets of identical twins and only 2 sets of fraternal twins. A whopping 89% of the twins in this group were identical twins.
The key to figuring out if gender dysphoria is genetic is to compare identical twins and fraternal twins. If identical twins are more likely to both have gender dysphoria than fraternal twins, you have a good case for a genetic contribution. So if the sets of identical twins are chosen in a different way from the sets of fraternal twins, you have a problem.
In fact, for this study most of the data on identical twins is coming from case reports; there might be a selection bias involved there. Almost all of the data on fraternal twins, however, is coming from comprehensive reviews of clinic records.
Reviews of case studies include a number of other problems.
You can’t be sure people are being diagnosed in the same way; you may not be looking at the same phenomenon. This review looked at people diagnosed by different therapists in at least ten different countries.* The dates of the studies ranged from 1956 to 2011. Some of the twins were children or teens, some were adults.
There may also be cultural or environmental differences that are relevant. For example, one of the case studies is of a pair of identical twins in Iran. Both twins are trans. We know that many people feel pressured to transition in Iran; what if that is a factor in this particular case? What if in another country, only one of the twins would have transitioned?
The era of the study might also affect gender dysphoria. For example, the Belgian clinic noticed that two of the fraternal twins with gender dysphoria they found had been born after in vitro fertilization. If IVF is a factor in gender dysphoria, it will only affect later cases.**
The data in case studies is not uniform; this makes it hard to compare. For example, one study discussed birth weights while another focused on relationships with parents.
This review of studies did include three sets of twins who were found in a method that did not have a selection bias or problems with inconsistent collection of data.
1) Zucker looked at the records of 561 patients who went to a Canadian clinic for gender dysphoria between 1976 and 2011 and found 25 sets of twins. The patients were all under 12 years old.
They found no cases where both twins had gender dysphoria.
2) Heylens and De Cuypere looked at 3 sets of adult twins from the 450 patients who went to a Belgian gender clinic between 1985-2011 plus 3 sets of non-adult twins who went to the Belgian gender clinic for children and teenagers.
They found only one case where both of the twins had gender dysphoria: a set of identical twins who were female-to-male transgender (FtM).
3) Vujovic et al reviewed all the cases of gender dysphoria who were treated at a Serbian clinic between 1987 and 2006. Out of 147 people, one trans man and one trans woman had a fraternal twin. Neither of their twins had gender dysphoria.
If we exclude case studies because of possible bias, we end up with no genetic component to gender dysphoria in trans women. None of the clinics found pairs of male twins who both had gender dysphoria.
The problem with this approach is that identical twins who are both male-to-female transsexuals exist. They just didn’t show up at these three clinics. Presumably, they are very rare.
Using the clinic studies for trans men we would have one set of identical Belgian twins who both had gender dysphoria, and one set of identical Canadian twins who did not. In addition, we would have three sets of fraternal twins where only one twin had gender dysphoria. This is not enough data.
So it makes sense to look at the data from individual case studies; we just need to be cautious about interpreting it. It is possible that it would over or underestimate the genetic component to gender dysphoria.
What was the data, then?
The authors searched the literature and put their data from the three clinics together with data from 17 different case reports and studies.***
FtMs with identical twins
3 sets of identical twins who both had gender dysphoria (37.5%)
5 sets of identical twins where only one of the twins had gender dysphoria (62.5%)
FtMs with fraternal twins
5 sets of fraternal twins where only one of the twins had gender dysphoria (100%)
MtFs with identical twins
6 sets of identical twins where both twins had gender dysphoria (40%)
9 sets of identical twins where only one twin had gender dysphoria (60%)
MtFs with fraternal twins
16 sets of fraternal twins where only one twin had gender dysphoria (100%)
Based on this data, identical twins with gender dysphoria are more likely than fraternal twins or the general public to have a twin with gender dysphoria. This suggests a genetic component to gender dysphoria.
However, most of the time, only one identical twin has gender dysphoria. This suggests other factors are involved in gender dysphoria.
At this point, we have no idea what the other factors involved might be. The case reports don’t give enough information on the twins to figure it out. The information they give is inconsistent; one study reported on the age of the first period while another talked about whether or not the mother was domineering. In addition, we may be comparing apples and oranges; for example, one study looked at an adult male American Indian in 1976, another looked at 13 year old American females in 1992.
The authors of the review conclude:
“The etiology of GID is a complex process of biopsychosocial components with unexplained interactions. Twin literature on GID supports the contribution of genetic factors to the development of gender identity with a higher tendency in males than in females.****
Since sample size is still limited and genotype studies are lacking, conclusions must be drawn with caution.
Therefore, detailed registers of GID twins, preferably on MZ twins discordant for GID and DZ twins are needed, to gain more decisive information about the influence of genetic vs. environmental factors in the development of GID.“
The authors of the study combine the data from studies of MtF and FtM twins for the statistical analysis. This gives them 9 pairs of identical twins where both twins had gender dysphoria (39%) and 14 pairs of identical twins where only one twin had gender dysphoria (61%). This is contrasted with 21 sets of fraternal twins where only one twin had gender dysphoria (100%). The difference is statistically significant.
This might be problematic since the mechanism that causes gender dysphoria in trans women is probably different from the mechanism that causes gender dysphoria in trans men. The genes are also probably different.
On the other hand the question here is whether or not gender dysphoria is inheirited, so perhaps this works.
Another problem is the possibility of selection bias. It looks like people are over-reporting cases involving identical twins. This might affect comparisons between identical twins and fraternal twins.
In addition, the total size of the group used in their statistical analysis is small and includes disparate groups – males and females, adults and children, people in different countries, and people living in different eras.
In the end, we’re left with weak evidence for a genetic component to gender dysphoria. We can’t prove it, but there is an excellent case for doing more studies in this area.
There is also an excellent case for future studies looking at what factors make one identical twin have gender dysphoria and one not. This seems to be the more common outcome than for both twins to have gender dysphoria.*****
Gender Identity Disorder in Twins: A Review of the Case Report Literature by Heylens G, De Cuypere G, Zucker KJ, Schelfaut C, Elaut E, Vanden Bossche H, De Baere E, T’Sjoen G in J Sex Med. 2012 Mar;9(3):751-7.
*Authors of the studies were from Belgium, Canada, Germany, Iran, Israel, the Netherlands, Serbia, Switzerland, the United Kingdom, and the United States. In addition, one author seemed to be from Scandinavia, possibly either Norway or Sweden.
**Using IVF means that the parents were infertile. It might be that the parents were older or that they had something wrong with their reproductive systems. It could be that the parent’s age or fertility problems affected the children rather than the IVF procedure.
***In some cases, it is hard to tell from the title if an article was a study or case report or an article on gender dysphoria that includes information on a case. Then again, the sub-title of this study is “A review of the case report literature,” so maybe they were all case reports.
**** I think the idea that there is a higher tendency in males than females is overstated. There were only 8 pairs of identical FtM twins and I doubt the 2.5% difference in the frequency of FtM versus MtF identical twin pairs who both have gender dysphoria is significant.
***** There may be selection bias in the cases of identical twins from the case reports. However, the clinic studies did find six sets of identical twins. In five of these six pairs, only one twin had gender dysphoria. Specifically, they found four pairs of identical male twins where only one twin had gender dysphoria, one pair of identical female twins where only one twin had gender dysphoria, and one pair of identical twins who were both FtM. So it looks like it is more common for only one identical twin to have gender dysphoria.
Emphases added are mine, including in the quote from the original review of the literature.