‘The idea that a person’s sex is determined by their anatomy at birth is not true, and we’ve known that it’s not true for decades.’ So says Dr Joshua D Safer, endocrinologist at the Center for Transgender Medicine and Surgery in New York and president of the United States Professional Association of Transgender Health.
There is a ‘significant, durable biological underpinning to gender identity,’ he says, adding: ‘As far as we in the mainstream biological-medical community understand it in 2018, it is hard-wired, it is biological, it is not entirely hormonal.’
Likewise the current US Endocrine Society Clinical Practice Guidelines from 2017 state that the term ‘biological sex’ is ‘imprecise and should be avoided’, citing the existence of intersex conditions and eliding them with transgenderism. The guidelines endorse the controversial notion that sex is ‘assigned’ not recognised at birth, and argue that ‘compelling studies’ show that biology contributes to ‘gender identity and its expression’.
Critics of transgender ideology often point to the fact that a person’s sex is biologically determined as male or female (XX or XY chromosomes) to demonstrate that any feelings a person has to be the other sex must be mistaken, and possibly a sign of mental illness.
While a tiny proportion (around 0.05 per cent) of people are genuinely ‘intersex’, meaning they exhibit a mixture of male and female biology owing to a disorder of sex development, transgenderism has typically been regarded as separate to this, as a psychological disorder known as gender dysphoria (and previously, gender identity disorder).
Increasingly, however, activists, including many within the medical establishment, are blurring the lines between the two and claiming that transgenderism, like intersexuality, is rooted in a person’s genetics and biology.
Such developments go some way to fulfilling the prediction made by noted transgender researcher Milton Diamond in 2013, that ‘transsexualism will eventually be seen as an intersex variation due to brain intersexuality’. ‘The evidence, I believe,’ he says, ‘is strong enough to consider transsexuality to be a form of brain intersex.’
One of the main sources of evidence for this striking claim is Diamond’s own twin study of 2013, the largest such study to date. In it he reviews all existing studies (involving 41 twin pairs) and adds a new survey of 69 twin pairs to give 110 pairs in total.
And the conclusion? That in 28 per cent of identical twin pairs in which one is transgender, the other one is also (known as a concordant pair). This compares to just 3 per cent (a single pair) of fraternal (non-identical) twins. The concordance rate rises to 33 per cent for male identical twin pairs.
This sounds impressive. Yet, as the American College of Pediatricians points out, in fact this confirms that gender identity is not genetically determined, as if it was the concordance rate would be approaching 100 per cent.
Still, 28 per cent (and 33 per cent for men) is not small and is the basis on which researchers like Diamond argue they have solid evidence that transgenderism is genetic and the result of a biologically ‘intersex brain’.
However, a close reading of Diamond’s study shows up the ways in which it greatly inflates claims about the proportion of double-transgender identical twins in the population.
Crucially, its sample of twins is in no way representative of the wider population and contains a hugely inflated number of concordant identical twin pairs.
Consider, for instance, its review of existing studies. There, 39 per cent (12 out of 31) of identical twin pairs are found to be concordant for transgenderism. Yet in the survey part of the study the corresponding figure is just 21 per cent (9 out of 43).
If the survey was representative this would mean the studies in the existing literature certainly were not. And indeed, the studies in the literature nowhere claim to be representative (they are mostly case studies), and it is hardly surprising to find researchers showing greater interest in concordant transgender pairs.
In addition, some of the ‘studies’ being relied upon here were actually, the paper explains, videos on Youtube and other websites, which are very likely to have a bias towards concordant pairs.
So should we take the survey as representative, and regard 21 per cent as the more accurate figure? Not so fast. The survey part of the study also makes no claim to be based on a representative sample, and it too shows heavy bias in favour of finding concordant pairs.
Take a look at how survey participants were found. The paper explains they were located through advertising for volunteers on ‘Internet websites’ and via ‘transsexual groups’ – a method that is surely likely to turn up a disproportionate number of concordant pairs, if only because there are two transgender individuals in the pair to spot and respond to the advert. There is also the important point that double-transgender identical twins are likely to be quite thrilled at being such and so more likely to respond to surveys.
Yet more evidence of bias is found in that three of the pairs of identical twins were raised separately – one separated at birth, one at the age of 4, and one at 14 – and all three were concordant, which is clearly not representative of separated twins in general.
This suggests that these pairs in particular were found because they were separated yet concordant, this being the feature that drew attention to them. Furthermore, the survey as a whole only turned up 9 concordant pairs, so the three raised separately make up a considerable proportion of the total (a third), skewing the results. It leaves us wondering how the other six were found.
As for the fraternal twins – where are they all? In the general population of the US and Europe fraternal twins outnumber identical twins by about ten to one (about 33 in 1,000 births are twins but only 3 in 1,000 births are identical twins, leaving 30 in 1,000 as fraternal twins). Yet in Diamond’s survey only 43 per cent (32 out of 75) of twin pairs are fraternal. It should be more like 90 per cent! Clearly fraternal twins were much less likely, for whatever reason, to participate in this survey – which will be one of the main reasons it did not register a single concordant fraternal pair.
Diamond argues at one point that his study, which he says combines ‘Internet survey data’ with ‘clinically reported material,’ can be taken as producing ‘representative data’. His main basis for making this bizarre and implausible claim is that his results are similar to a review article from 2012 which found 39 per cent (9 out of 23) of identical twin pairs concordant compared to zero (out of 21) fraternal twin pairs. Yet 39 per cent is nowhere near the 28 per cent concordance rate in his study.
And while it does match the 39 per cent concordance rate in his own literature review, all that surely shows is they were drawing on much of the same literature. What it doesn’t do is magically transform the cases in the literature into a representative sample.
So 39 per cent is not representative, neither is 21 per cent (from Diamond’s survey), and combining the two to give 28 per cent is certainly not representative either. You only get a representative sample by generating participants randomly and ensuring nothing skews the selection of participants, not by sticking together two unrepresentative samples.
This is not to say genes play no role in a person’s susceptibility to transgenderism and gender dysphoria – many illnesses, including mental illnesses, run in families as a result of shared genes.
But researchers like Diamond appear to be keen to play up the role of genes and biology to support the notion of an innate gender that takes precedence over the sex of the mere body. This seems to include being happy to give the impression that a study with a heavily biased sample is a fair reflection of the wider population when it is nothing of the sort.
What, then, is the true figure? The truth is, we still have no idea, though it is unlikely to be anything like 21 per cent. 5 per cent? 1 per cent? More research is needed that eliminates the sample biases if we are to have any chance of finding out.
But however this turns out, it’s important to recognise that just because there is a genetic component to transgenderism, that doesn’t make it healthy. If a person thinks they are a different sex to their body and chromosomes then it remains the soundest response to regard the problem as residing in their psychology and treating it there, even if it has arisen in part from some underlying biological factors.
Transgender proponents are enjoying great success at the moment promoting the novel idea that it is the body that is faulty when the mind thinks differently. Yet this is a dangerous notion that leads to all kinds of problems, from confused and mutilated children to exposing women to increased risk of assault and undermining their sporting achievements.
Studies like Diamond’s, which exaggerate the links with biology and genes by passing off skewed samples as representative data, constitute the supposedly scientific underpinning of these efforts. Their errors and misrepresentations must be illuminated and exposed.