Trans Brains: The Neurological Basis of Gender Identity

Decades of neuroimaging and postmortem research have identified real, measurable differences in the brains of transgender individuals compared with cisgender people of the same birth-assigned sex. These differences show up in the size of specific brain nuclei, the thickness of the cortex, the microstructure of white-matter tracts, and the way functional networks communicate. None of these findings point to a single “gender identity center” in the brain, and the field remains far from a diagnostic scan. But the accumulated evidence does suggest that gender identity has roots in how the brain develops, shaped by hormones, genes, and epigenetic factors that can push neural development in a direction independent of genital anatomy.

The Postmortem Studies That Started It All

The earliest hard evidence came from examining brains after death. In 1995, a Dutch team reported that a small region of the hypothalamus called the central subdivision of the bed nucleus of the stria terminalis (BSTc) was larger in men than in women, and that in transgender women, its volume was female-sized. The finding was independent of sexual orientation, and adult hormone levels did not appear to account for it.¹ Five years later, the same group counted the neurons in that region and found that men had roughly twice as many somatostatin-producing neurons as women. Transgender women had neuron counts in the female range, and a transgender man’s count fell in the male range. Again, hormone treatment during adulthood did not seem to explain the pattern.²

A later postmortem study examined a different hypothalamic cluster, INAH3, and found the same general story. Transgender women had a volume and neuron count similar to cisgender women, while a transgender man’s INAH3 was in the male control range, even though testosterone treatment had been stopped three years before death.³ These postmortem studies are powerful because they provide direct cellular-level data, but they are based on very small samples, sometimes just a handful of brains. They opened the door, but they could not tell us about the living brain at scale.

Cortical Thickness and Gray Matter

Modern MRI studies allow researchers to measure the thickness of the cortex and the volume of gray matter in living people. The results paint a picture that is more nuanced than “transgender people have the brain of their identified gender.” One study of transgender women who had not yet started hormone therapy found that their cortical thickness did not differ from cisgender women in any region, while it was thicker than cisgender men across a wide swath of areas including the orbitofrontal cortex, insula, and medial occipital regions. Transgender men, similarly, showed cortical thickness that did not differ from cisgender women but was thicker than cisgender men in the right parietal lobe and medial parieto-temporal regions.⁴

Another study reported thicker cortex in transgender women compared with cisgender men in widespread bilateral regions, with no region where cisgender men had thicker cortex than transgender women.⁵ Gray matter volume tells a slightly different story. A voxel-based study found that both treatment-naïve and hormone-treated transgender women had lower gray matter volume in the bilateral insula compared with cisgender women.⁶ The overall picture from cortical research is that transgender brains are not a clean match to either sex category. Instead, they seem to present a distinct pattern that draws from both but also diverges from both.

A comprehensive review put it this way: the brains of transgender women show a complex mixture of masculine, feminine, and “demasculinized” regions, while transgender men show feminine, masculine, and “defeminized” regions. Neither pattern matches the typical brain profile of cisgender men or cisgender women.⁷

White Matter Microstructure

White matter is the brain’s wiring: long fibers coated in myelin that carry signals between regions. Diffusion tensor imaging (DTI) can measure how organized that wiring is, using a metric called fractional anisotropy (FA). Higher FA generally means more tightly bundled, more structurally coherent tracts. One early DTI study of transgender men before any hormone therapy found that their FA values in several major tracts were higher than cisgender women’s and closer to cisgender men’s, particularly in the posterior portion of a long bundle connecting frontal and parietal regions.⁸

A later study that carefully controlled for sexual orientation found a more specific pattern. After accounting for orientation, transgender participants showed lower, sex-atypical FA in two specific tracts: the right inferior fronto-occipital fasciculus and the left inferior longitudinal fasciculus. In all other tracts, their FA values became sex-typical once orientation was controlled.⁹ This finding highlights how tangled the research can get: some apparent brain differences in transgender individuals may partly reflect sexual orientation rather than gender identity per se, and separating these two dimensions requires careful study design.

One well-powered DTI study found no significant differences in FA maps between transgender and cisgender groups at all.¹⁰ This inconsistency is a recurring theme in the field. Sample sizes are small, scanning protocols differ, and whether researchers account for confounders like sexual orientation or overall brain size can change the result.

Functional Networks at Rest

Structure is only half the story. How brain regions talk to each other when a person is simply lying in the scanner doing nothing (resting-state functional connectivity) reveals something about the brain’s baseline organization. A study of transgender women found they had weaker functional connections than cisgender people in circuits involving the ventral attention and sensorimotor networks. Their local network efficiency was also lower.¹¹ In transgender men, weaker connectivity was found within the salience network and between the salience network and the default mode, executive control, and sensorimotor networks.¹²

A study using an “intrinsic ignition” framework, which measures how well one brain region can trigger a cascade of activity in other regions, found that this metric differentiated all four groups (cisgender men, cisgender women, transgender men, and transgender women) across the whole brain. Cisgender men showed higher ignition in attention and executive control networks than the other groups, while cisgender women showed higher ignition in the default mode network compared with both cisgender and transgender men.¹³ What is striking is that transgender individuals consistently landed in distinct functional territory, not simply matching their identified gender or their birth-assigned sex.

Why the Brain and Genitals Can Diverge

How can the same person’s body develop as one sex while key brain structures develop along a different trajectory? The answer appears to lie in timing. Genital differentiation occurs in the first two months of pregnancy, while sexual differentiation of the brain begins in the second half. Because these two processes happen during different windows, they can be influenced independently by fluctuations in hormones, stress, or other prenatal factors.¹⁴ This means an ambiguous genital outcome at birth does not necessarily reflect the same degree of brain masculinization or feminization.¹⁵

Animal experiments have confirmed this concept directly. In sheep, exposure to testosterone from day 30 to day 60 of gestation masculinized the genitals but left a key sexually dimorphic brain nucleus unchanged. Exposure from day 60 to day 90 did the opposite: genitals were normal but the brain nucleus was enlarged to the male-typical size.¹⁶ These are separate critical windows that can be experimentally uncoupled, providing a biological mechanism by which genital sex and brain sex can end up misaligned. Rodent models have similarly shown that steroid hormones during early development shape processes like neurogenesis, cell migration, and synaptic patterning in brain regions that later show sex differences.¹⁷

Genetic and Epigenetic Clues

If prenatal hormones are central, genetics shapes how the body responds to those hormones. A large candidate-gene study of transgender women found significant associations with variants in genes involved in sex hormone signaling. Having a specific genotype in a gene for a sulfotransferase enzyme (SULT2A1) increased the likelihood of being transgender by about 1.6 times, and a variant in the estrogen receptor alpha gene (ERα) increased it by a similar magnitude.¹⁸ These are modest effect sizes, and no single gene variant comes close to “causing” a transgender identity. But they suggest the hormonal pathway is subtly tuned differently in some transgender individuals at the DNA level.

Epigenetics, the chemical modifications that sit on top of DNA and influence which genes get read, is an emerging piece of the puzzle. An epigenome-wide study found that cisgender and transgender individuals differed in global patterns of DNA methylation even before any hormone therapy. Among people assigned male at birth, dozens of specific sites showed meaningfully different methylation between cisgender and transgender participants.¹⁹ These epigenetic differences could reflect prenatal hormone exposures that left lasting marks on gene regulation, though the research is still early and the causal direction is uncertain.

Hormone therapy itself also shifts the epigenetic landscape. A study of the estrogen receptor gene ESR1 found that before treatment, transgender men had the lowest methylation levels compared with all other groups, while transgender women had intermediate levels between cisgender men and cisgender women. After hormone therapy, both transgender groups converged on an intermediate methylation level.²⁰ Another study confirmed that gender-affirming hormone therapy changes thousands of methylation sites in the blood, but methylation patterns established in early development on the X chromosome and sex-specific autosomal sites are largely resistant to change, with only about 3% affected.²¹ In other words, hormones can rewrite some of the epigenetic script but not all of it, and the earliest marks prove the most durable.

How Hormone Therapy Reshapes the Brain

Gender-affirming hormone therapy does not just change the body; it measurably alters brain tissue. A neuroimaging study tracking changes over roughly four and a half months found significant shifts in gray matter density and microstructure across multiple regions, including the insula, cingulate cortex, fusiform gyrus, and cerebellum.²² Longitudinal work on the hypothalamus found that in transgender women, total hypothalamic volume decreased over time during estrogen-based therapy, with changes concentrated in several specific subunits. These decreases differed significantly from the stable volumes seen in cisgender controls.²³

A preprint tracking cortical thickness changes over time reported that hormone therapy drove the cortex in opposite directions in the two groups: localized thickening in transgender men and widespread thinning in transgender women, with both converging in the occipital cortex. The spatial patterns of these changes matched the known distribution of sex hormone receptors across the brain, and improved psychological well-being tracked alongside the neurobiological shifts.²⁴ Serotonin systems respond too: high-dose testosterone therapy in transgender men increased serotonin transporter binding, which has implications for mood regulation.²⁵ Together, these findings show the adult brain remains remarkably responsive to sex hormones, but they also raise the question of how to interpret pre-treatment brain differences when post-treatment brains change so dramatically.

The Body Map in the Brain

One of the most personally resonant aspects of being transgender is the feeling that certain body parts do not belong. There is emerging evidence that this disconnect has a neural basis. A study using magnetoencephalography found that when transgender men were given a tactile stimulus on the chest, a body part most described as feeling incongruent, their brains showed significantly reduced activation in the supramarginal gyrus and secondary somatosensory cortex compared with cisgender women. At the same time, activation was increased at the temporal pole. The white matter in these regions was also structurally different.²⁶ The implication is that the brain’s map of the body may literally underrepresent the incongruent part.

Resting-state connectivity data supports this idea from a different angle. Transgender men showed decreased connectivity in brain regions associated with self-referential processing (the default mode network) and weaker connectivity in visual networks involved in body perception. When they viewed digitally morphed bodies that matched their identified gender, stronger activity in the anterior cingulate within the default mode network predicted higher “self” ratings, suggesting that their self-identification is linked to reflective processing rather than the more automatic, gut-level body recognition seen in cisgender controls.²⁷

The Mosaic Problem

All of the above research exists in tension with a broader finding in neuroscience: human brains are not neatly “male” or “female.” An analysis of over 1,400 brain scans found extensive overlap between males and females on every measure of gray matter, white matter, and connectivity examined. Brains that were internally consistent, with all features falling at the “male” or “female” end, were rare. Most brains were unique mosaics, combining features more common in one sex with features more common in the other.²⁸ This mosaic nature held up across datasets, ages, and analysis methods, and paralleled a similar mosaic pattern found in personality traits, interests, and behaviors across thousands of people.²⁹

This finding does not invalidate the research on transgender brains, but it does complicate the framing. Saying transgender women “have female brains” is an oversimplification. What the data more accurately shows is that transgender individuals have particular constellations of brain features that differ from cisgender people of their birth-assigned sex, often (but not always) in the direction of their identified gender. The differences are real, but they sit within a much broader landscape of individual variation where almost nobody has a brain that is 100% “male-type” or “female-type” on every dimension.

Where the Research Runs Thin

One glaring gap is the near-total absence of non-binary people from neuroimaging research. A scoping review of brain studies on gender-diverse youth noted that non-binary identities were “mostly not taken into account.”³⁰ A population-level study of adolescents using data from thousands of participants found that those who reported gender diversity did not differ in overall brain volume from those who did not. Among those assigned female at birth, gender-diverse adolescents showed slightly smaller thalamic volumes, but this finding did not survive correction for multiple comparisons.³¹ That study’s honest conclusion is that we simply do not have strong evidence yet for brain structural correlates of gender diversity in adolescents at the population level.

Sample size is a persistent problem throughout the field. Most neuroimaging studies involve a few dozen participants per group, and some of the most cited postmortem studies involved fewer than ten transgender brains. Statistical power is low, and the signal-to-noise ratio in brain scans of any rare population makes it difficult to separate true effects from noise.³² The field has also been heavily tilted: a historical review found that most neuroscience research on gender and sexual diversity centered on gay men (about a third of studies) or transgender people (about a quarter), with very little work on lesbian, bisexual, or other identities.³³

Cognition and Hemisphere Asymmetry

A question people sometimes ask is whether transgender individuals “think” more like their identified gender. One study tested mental rotation (a spatial task where men on average score higher) and lexical decision-making (a verbal task) in cisgender men, cisgender women, and transgender men. In mental rotation, transgender men performed at the same level as cisgender men, both outperforming cisgender women, but this group difference disappeared once general intelligence was factored in. In the verbal task, though, transgender men and cisgender men both showed a stronger left-hemisphere advantage than cisgender women, and this difference was not explained by intelligence. The researchers interpreted this as a hemisphere-asymmetry effect moderated by gender identity itself, not by birth-assigned sex.³⁴

The Intersection with Autism

Clinicians have long observed that transgender and gender-diverse populations include a higher proportion of autistic individuals than the general population. Brain research is beginning to explore this overlap. One study of transgender youth found significant autism-related differences in default mode network connectivity that tracked with diagnostic indicators of autism, not with levels of gender dysphoria, internalizing symptoms, or stigma related to being a sexual or gender minority.³⁵ In other words, the autism-related brain patterns appeared to be genuine developmental autism, not an artifact of being transgender or of the psychological distress that sometimes accompanies gender incongruence. A community-academic dialogue on this intersection noted that autistic transgender people show brain phenotypes consistent with cisgender autistic people, while their implicit gender patterns are consistent with non-autistic transgender people.³⁶ The two conditions appear to co-occur without one mimicking the other in the brain.

Puberty Blockers and the Developing Brain

One of the most politically charged questions in this area is what puberty-blocking medications (GnRH agonists) do to the adolescent brain. The honest answer is that the evidence is very thin. A study comparing treated and untreated transgender youth on an executive function task found no group differences in brain activation patterns and no significant performance differences, with one exception: treated transgender girls performed worse than untreated transgender boys and cisgender controls on the task itself.³⁷ A case study following a single transgender girl over 28 months of puberty suppression found that white matter maturation did not progress as expected during that time, and working memory scores declined. A consensus parameter review noted these two studies and concluded that larger, longer-term longitudinal research is needed to understand the neurodevelopmental effects of puberty suppression.³⁸ Anyone making confident claims in either direction about cognitive effects of puberty blockers is outrunning the data.