Pervasive misrepresentation of twin evidence for a genetic component of gender dysphoria

Major voices in the contemporary anti-trans movement, including American College of Pediatricians president Dr. Michelle Cretella and Heritage Foundation research fellow Ryan T. Anderson, have recently put forth questionable claims about studies of gender dysphoria in twins. Last year, Michelle Cretella argued:

But in the largest study of twin transgender adults, published by Dr. Milton Diamond in 2013, only 28 percent of the identical twins both identified as transgender. Seventy-two percent of the time, they differed. (Diamond’s study reported 20 percent identifying as transgender, but his actual data demonstrate a 28 percent figure, as I note here in footnote 19.)

That 28 percent of identical twins both identified as transgender suggests a minimal biological predisposition, which means transgenderism will not manifest itself without outside nonbiological factors also impacting the individual during his lifetime.

And in his noxiously titled book “When Harry Became Sally”, Ryan T. Anderson claims:

There is no scientific evidence that a transgender identity is biologically determined. In fact, there is some evidence—though by no means conclusive—that other factors must be involved. Professor Paul Hruz cites studies on identical twins, with the same genetic complements and the same prenatal environment, who developed differing gender identities. If transgender identity were innate and independent of nurture, then two children who shared a womb and have identical genetic material would both be transgender, or neither would be. But that isn’t what the research indicates.

For example, a study by Milton Diamond published in the International Journal of Transgenderism looked at transgender individuals who have an identical twin, and found that the twin also identified as transgender in 20 percent of the cases. That figure suggests some measure of biological predisposition, but far from biological determination. Dr. Quentin Van Meter noted in court testimony that the concordance rate would be close to 100 percent if genes and/or prenatal hormones were the primary causes of transgender identities. Clearly, life experiences play a large role. Neither twin studies nor brain imaging supports the hypothesis that a transgender identity is innate and fixed, or that a person could really be “a man trapped in a woman’s body” or “a woman trapped in a man’s body.”

Cretella and Anderson’s argument misleadingly portrays the evidence from twin studies and selectively omits relevant information. Their assertion that biological predisposition to gender dysphoria is negligible because identical twins are not universally both dysphoric or non-dysphoric gives the false impression that a finding of concordance in 20% or more of identical twins is somehow spurious or meaningless. It is not.

What Cretella and Anderson neglected to inform their readers is that concordance for gender dysphoria among dizygotic (fraternal) twins is far more rare. A 2012 review of the literature found that of the reported cases of identical twins where at least one was gender dysphoric, 39.1% of twin pairs were concordant for gender dysphoria (Heylens et al., 2012). However, out of all the reported cases of fraternal twins where at least one was gender dysphoric, none were concordant for gender dysphoria. In other words, genetically identical twins were vastly more likely to be concordant for gender dysphoria than those who were not genetically identical. That is not a negligible biological contribution to gender identity – it is a significant one.

Additionally, Cretella and Anderson’s assertion that a lack of concordance in some identical twins points to a lack of a biological determination for gender dysphoria is faulty. There are numerous possible biological influences on gender identity other than the genetic code itself, as Heylens et al. (2012) explain:

The higher concordance for GID in MZ than in DZ twins is consistent with a genetic influence on its genesis although shared and nonshared environmental factors cannot be ruled out. … The discordance for GID in five (62.5%) out of eight of the described MZ female twin pairs might have several explanations. For example, differential prenatal and postnatal environmental factors might both contribute to discordance between MZ twins. …

In addition, several epigenetic differences may underlie phenotypic discordance between the MZ twin pairs. In female MZ twins, skewed X inactivation in the affected twin might be an illustration of an epigenetic difference explaining phenotypic discordance. A study by Fraga et al. demonstrated that epigenetic differences increase during the lifetime of MZ twins, which may help to understand phenotypic differences. Genetic differences between concordant and discordant MZ twins are illustrated in a recent study by Bruder et al., showing copy number variations (CNVs) in their genomes. These findings suggested that CNV analysis in phenotypically discordant MZ twins may provide a powerful tool for identifying disease predisposition loci.

On the basis of these and other twin and family studies, Polderman et al. (2018) have proposed a polygenic threshold model for gender identity, concluding that there is “significant and consistent evidence for the role of innate genetic factors in the development of both cisgender and transgender identities, a negligible role for shared environmental factors, and a small potential role for unique environmental factors.” Cretella and Anderson do a disservice to their readers when they neglect this evidence in favor of their own motivated oversimplifications of “biology”. ■