What We Know About Trans Brains

Understanding scientific research and dispelling a community myth

Author’s Note: The Grinding Glass series will resume shortly. The end of a semester is a busy time, and articles in that series are unusually time-consuming to write.

---

I sigh, tiredly.

Yet again, someone, somewhere has posted a confident, categorical statement somewhere on social media (is it on Mastodon? Reddit? Bluesky? TikTok? Dear god, is it Facebook? Please tell me it’s not Facebook—I don’t have the heart for Facebook anymore) about how trans brains are more like the brains of our lived gender than our gender assigned at birth. Or… or is it progesterone’s effects on breast development this time? Or maybe the interactions of testosterone and finasteride on transmasculine bottom development? Or how spironolactone is somehow dangerous? Or, could it be, something about specialty, high-dose topical minoxidil for hair restoration?

No. No, this time it’s certainly trans brains, and not one of the dozens of other pieces of lore that floats around trans communities, almost entirely uncontested and equally unsupported by anything even resembling robust medical research. And really, I understand. I do. There are such huge gaps—gulfs, really—in our scientific knowledge about trans bodies and gender-affirming medicine, and trans people are so overwhelmingly attacked on every side by transphobic pseudoscience… We need things to believe in. A sense of control over a world that seems like it’s spinning out of control on so many levels.

Hope that, finally, our bodies will be right. That we can take concrete, specific action to make them right.

I roll my shoulders yet again, call up the same old article on trans brains, and write out the same comment for fiftieth (hundredth?) time. Yes, I say, you’ve got ten studies there that say trans brains are shaped like the brains of our lived gender. No, none of them turned out to be true. This one study here not only refutes them all, it does so by using their own data—and the data of hundreds of other studies that found a myriad of other little differences—to do so. No, you can’t rebut the source with another ten sources. They’re all part of my study’s dataset.

Your articles don’t say what you think they say.

Scientific research writing doesn’t work the way you think it does.

Post. The little missive flutters away into the screaming churn if the internet, a billion voices screaming at another billion voices screaming at—

Oh, look. Another post about trans brains. I resist the urge to just copy-and-paste my last reply, because this one’s coming at it a different way. I fight not to respond with a curt rebuttal. Everybody deserves a chance to grow and be treated with kindness and gentleness. I write, again, as patiently as I can manage.

Post.

Oh, look. Another—

Eggs, trivia, and other scientific bullshit

I know, you’ve heard about trans brains being built like those of our lived gender. It’s probably upsettling to hear it all brushed aside so casually. I promise—I’ll get to it. I promise.

But before we can talk about the what, we need to talk about the why, and the how. That’s the boring part of science. Methodology sections are always the worst to slog through, but they’re the most important by far. Here’s the upside, though: I’m going to make this into one of those neato moments where I pull the curtain back and show you how Stained Glass Woman gets made!

The first thing you need to understand is that scientific progress isn’t made in leaps and bounds. It’s made by thousands of researchers writing hundreds of articles on very, very specific, very, very small things. Each article builds on the findings of the ones that came before it until, eventually, the whole body of research passes a threshold and we’re finally able to say Something Big.

But along the way? Along the way, we publish some really trivial crap.

And that really trivial crap? What you hear about it isn’t even what that really trivial crap even said.

And that’s okay! In scientific research, finding out “nope, there’s nothing here” is exactly as important as finding “yep, there’s something here!” But the news is a hungry beast, so it makes a big deal out of those little bitty research findings even when there really isn’t a big deal to be made. And that, by the way, is why it seems like medical research can’t make up its mind whether eggs are good, bad, neither, or both.

When I write Stained Glass Woman, I do my best to avoid this sort of thing, and stick to research that’s either done with a high level of confidence, or include and explain all of the asterisks from the original research. Take the article I wrote on DMRT-1, for instance: there’s a whole section on why trying to do this now would definitely kill you super dead, and how we need whole new types of technology before we could even seriously look at trying this sort of thing on a living person.

You’ve definitely picked up a smattering of scientific myths, both trans and not—there’s just too many out there to not—along the way. That’s normal. But when we’re trying to learn and understand what we actually know, in a scientific sense, it means we need to do some legwork.

Understanding scientific research writing

So, if you go back to that first comic, there’s one type of paper on there that really matters: the one titled “We’re 500 scientists and here’s what we’ve been up to for the last 10 years.” It’s parodying some typical big releases for physics research, and if you think it’s an exaggeration, this is the full author list for one of six big releases that came out last summer, when astrophysicists were able to prove the existence of gravitational waves (this is apparently a big deal in astrophysics, but I’m not an astrophysicist, so hell if I know):

And if you’re like, “well, 69 is less than 500,” remember that “The NANOGrav Collaboration” is author #69. So, a whole collection of people who helped, but not enough to get to put their name on the article.

And this is just one of the six articles that were released as part of a package.

So yeah. Not exactly an exaggeration.

This was what is known as a large randomized control trial, and it’s an example of one of the best kinds of research we can do. In a nutshell, an RCT tests a group (with a bunch of controls) of whatever it is that they want to test. They randomly split a bunch of members of that group into two subgroups, which are as identical as possible. Then, one picked at random is tested with the hypothesis that they’re testing, the other one gets no intervention, and we watch both groups equally to see what happens. However—and this is critical—not even the researchers know which group is which! In drug research, for instance, they manage this by making a sugar pill (or injection or whatever) that looks identical to the pill that the researchers are actually testing, and only the person preparing the medicine for the trial, who isn’t part of the study, and doesn’t talk to anyone on the team, knows which is which (and even then, only by a patient identifier number, never by name).

We do this because when we do research, we want our hypothesis to be true. After all, we’re putting all this time and effort into trying to prove it! But that means that researchers can subconsciously put their thumbs on the scale, so to speak, if they know who’s in which group.

The awesome thing about an RCT is that if you get a good result out of a big (because the bigger the RCT is, the more reliable it is) RCT, that’s a really strong step towards saying that yes, this thing is true, or no, this thing is not true.

But the thing is? RCT’s aren’t even the best research we can do.

Meta-analyses! Heavenly choir sings

RCT’s are only a very strong way of studying something. The very best way we have of making scientific progress is called a meta-analysis.

The problem with RCT’s is that, eventually, you’ll have a bunch of them, and not all of them will say the same thing. Say, for instance, you have a very convenient bibliography of 72 studies on the effects of transition on transgender well-being, and it’s a good mix of RCT’s, cohort studies, and case-control studies. You look at them all, and you find that 93% of those studies found that transition improves trans well-being, while 7% either had mixed (both positive and negative) findings or found no result. I mean, 93% is a really strong number, right? But why do we get so high, but not to 100%? Seems… kind of odd, doesn’t it? And what about the research papers that transphobes wave around, that say things somehow get worse for trans people after transition (like that one Swedish study that the author has been trying to keep people from misrepresenting since the day she published it)?

Well, science and biology are messy. Human error creeps in, and there really isn’t anything at all that has to do with human beings that applies to everyone everywhere; we’re just too diverse. Then, on top of that, people with political agendas like to hijack our research and claim it says stuff that it really, really, really doesn’t.

What if we could take all of those studies, with all the different types and methods—RCTs and cohort studies and case-control studies—that looked at slightly different things in slightly different ways and make them talk to each other, in a scientific sense.

We can. That’s what a meta-analysis is, and if you’re sitting there reading this and thinking to yourself that that sounds like it’s incredibly hard to do, you’re exactly right. Putting a good meta-analysis together is a monumental task that takes years and years of work. Like any good study, the bigger the sample size, the better, but since every study you add means you need to tweak how all the other studies talk to each other, the problem gets bigger and bigger as your body of research data grows.

But when we do manage it, the results can be incredible.

And that’s where we needed to get to before we could talk about trans brains.

Your brain isn’t different because you’re trans

One of the most persistent myths in our entire community is that a person is trans because our brain structures are more like our real gender than our gender assigned at birth. In many ways, this hypothesis has been the transgender version of the old 00’s hunt for the gay gene (spoiler, there isn’t one; it’s complex genetic and epigenetic influence, just like being trans is)—it’s given us a biological explanation for why we’re trans, and why gender transition is an absolute necessity for so many of us. It’s deeply seductive, because, if true, there’s obviously no way to magically change our brains. It’d mean that anyone denying who we were was just plain scientifically illiterate on a really, really basic level.

It’s a shame it isn’t true. Would’ve been awfully convenient if it were.

That hypothesis was dealt an absolute death-blow a few years back, though, and it wasn’t even the direct target of the death-blow. Eliot et al, in 2020, published the largest meta-study ever even attempted on the structure of human brains and how they were related (or not) to sex, including all research done in the last thirty years on brain structure. They weren’t even looking at trans brains here, not directly—though they included all of the research on trans brains too, since it met their inclusion criteria—they were looking at how gender and sex interacted for brain structure for all people, trans and cis. After all this work, they found that there was no significant difference between male and female brains so decisively that they titled the damn paper “Dump the ‘dimorphism.’” Their research said, in short, that only about 1% of the brain had any difference at all between male and female people, structurally, and the only real difference was sheer size, because males are, on average, bigger than females. And, while I could go on to explain how this all is the case point by point, I think Eliot et al’s conclusion really says it best:

The term “dimorphism” reinforces a binary understanding of [sex/gender] brain difference, when in fact, few such differences actually exist and the ones that do are very small, with great variability from population to population. Of note, a similar conclusion is converging from the growing research on transgender participants’ brains, according to Smith et al. (2015) who remark that “viewing gender as a binary or dichotomous category has to be reconsidered.”

Rather, a picture is emerging not of two brain types nor even a continuous gradient from masculine to feminine, but of a multidimensional “mosaic” of countless brain attributes that differ in unique patterns across all individuals (Joel et al., 2015). Although such differences may, in a particular sample, sum up to discriminate male from female brains, the precise discriminators do not translate across populations.

And if there’s no such thing as a male brain or a female brain, a transfeminine person’s brain cannot be more like a female brain, nor a transmasculine person’s brain more like a male brain. It makes sense, too—with the old gender-sex brain hypothesis, what would an agender person’s brain look like, if it were true? A bigender person’s? The whole hypothesis only works if we ignore the fact that nonbinary people exist.

For clarity: we definitely shouldn’t ignore the fact that nonbinary people exist.

Now, if you root around in PubMed, which is a lot of what I do to prep an article for Stained Glass Woman, you’ll find conflicting stuff. Say, for instance, this mega-analysis (a type of meta-analysis) that finds notable differences between transgender and cisgender brains, or this systematic review (another type of meta-analysis) which struggled to do so, both published after Eliot et al’s work. What do you do when you have a clash of these top-level sources?

Look at the methods. Yeah, it’s boring. It’s what matters, though.

Our example mega-analysis here includes a mere 803 participants, smaller than most RCT’s. That systematic review includes only 63 articles, whereas Eliot et al reviewed hundreds. When it comes to sample sizes, bigger is always, always better. Why?

Because biology is messy, and there’s no such thing as a universal rule for human beings or bodies. That means that when you do research, you’re gonna have oddball results from time to time. Outliers are normal—good—parts of human research.

But you can’t hold them up over better research just because you like what they say.

Community lore

One of the big problems in the trans community is just how much research hasn’t been done, and in the absence of real and meaningful research, myth and lore abound. We cling to hope and seek out whispers of what-might-be, what might better develop our bodies in the way we want. I’m as guilty of this as anyone else—when I didn’t get much top growth after a year on estrogen, I dove for progesterone, just like almost every other transfeminine person in my position does. All this despite the fact that there’s no good research to support progesterone having a role in breast development or maturation and more than a little that cast doubt on it having any such role in cis or trans women.

I knew, at the time, that the research said it wouldn’t help. I did it anyway.

Hope does that to a person.

And yeah, you hear all the time from transfeminine people that they saw their boobs grow when they started progesterone. That doesn’t mean it played a role. Spurious correlations are a thing.

Now, none of this means that progesterone isn’t crucially important for transfeminine people! For instance, it plays one half of a really important dual role with estrogen in forming and strengthening our bones, and is key to preventing dementia over the long term.

Just… probably not boobs. We can’t say for certain right now. But probably not.

What all this does mean is that we need to remember, when we try to learn from research done on us and about us, that we have a stake in the findings of that research. Just like scientists can subconsciously affect the outcome of RCT’s if they know who’s in the control group and who’s in the test group, what we want to be true affects what we believe to be true.

That’s not good science.

And I believe that trans folks deserve the benefit of the best, most recent, and most useful research done about us. Sometimes, though, that means letting go of the things we want to be true…

But aren’t.