Why Are People Transgender? What the Science Shows

Being transgender appears to stem from a mix of biological factors, especially genetics and the hormonal environment in the womb, rather than from any single gene, parenting style, or social influence. Research from genetics, endocrinology, and brain imaging all converge on the same broad finding: gender identity has deep roots in biology, even though no one has pinpointed a single “cause.” The Endocrine Society, one of the world’s leading medical organizations on hormones and development, summarizes the evidence by stating that gender identity likely reflects “a complex interplay of biological, environmental, and cultural factors.”¹ That interplay is worth unpacking, because the individual lines of evidence are more interesting than a single-sentence summary can convey.

Genetics Play a Substantial Role

One of the strongest tools researchers have for teasing apart nature from nurture is twin studies. Identical twins share nearly all of their DNA, while fraternal twins share roughly half, the same as any siblings. If a trait is partly genetic, you’d expect identical twins to match on that trait more often than fraternal twins do. That is exactly what researchers find with gender identity.

A pooled analysis of 463 twin pairs found that about 21% of identical twin pairs were concordant for being transgender, compared to roughly 9% of fraternal pairs. When the researchers calculated relative risk using a 1% population prevalence estimate, the numbers were striking: identical twins were about 21 times more likely than the general population to both be transgender, while fraternal twins were about 9 times more likely.² The gap between those two numbers points to a substantial genetic contribution. If genes played no role, identical and fraternal twins would show similar rates. If genes were the entire story, identical twin concordance would be close to 100%. Instead, the data land somewhere in between: genes matter a lot, but they are not the whole picture.

That “not the whole picture” part is important. The fact that most identical twins of transgender individuals are not themselves transgender tells us that shared DNA alone is not sufficient. Something else in the mix, whether prenatal hormones, epigenetic switches, or other developmental factors, helps determine the outcome. This is consistent with how many complex human traits work: height, temperament, and even handedness all have strong genetic components without being purely genetic.

Hormones in the Womb Shape Gender Identity

The leading biological theory focuses on what happens before birth. During fetal development, the brain and the body respond to sex hormones at different times and in somewhat independent ways. The body’s reproductive anatomy forms early in pregnancy, while the brain regions associated with gender identity develop later. If the hormonal signals during those two windows differ, even subtly, the result can be a person whose brain-based sense of gender does not align with their body’s anatomy.

A review of evidence across disciplines found that prenatal testosterone exposure plays a key organizing role. Gender identity tends to shift in a more masculine direction when the developing brain is exposed to higher testosterone levels, and in a more feminine direction when it is not. This conclusion draws on animal research, clinical observations of people with differences in sexual development, and biometric indicators of androgen exposure.³

Some molecular evidence adds texture to this picture. The androgen receptor gene contains a stretch of repeated DNA that varies in length from person to person. Longer repeats are associated with weaker sensitivity to testosterone. A meta-analysis of case-control studies found that transgender women (people assigned male at birth who identify as women) tend to have slightly longer repeats in this gene compared to cisgender men.⁴ In plain terms, their bodies may respond a bit less strongly to testosterone during critical periods of brain development. The effect size is small, and this one gene variant certainly does not explain everything. But it fits neatly into the broader hormonal theory: even modest differences in how the brain processes sex hormones during development could shift the outcome.

Brain Structure Differences Exist Before Hormone Therapy

A recurring question is whether transgender people’s brains look different from cisgender people’s brains, and if so, whether those differences are innate or a consequence of hormone therapy. The largest neuroimaging study on this topic to date, a mega-analysis from the ENIGMA Transgender Persons Working Group, addressed that question by specifically studying people who had not yet received any hormonal treatment. The study analyzed structural brain scans from more than 800 participants, including transgender men, transgender women, cisgender men, and cisgender women.⁵

The findings did not reveal a simple picture where, say, transgender women’s brains look just like cisgender women’s brains. Instead, several brain measurements in transgender individuals fell between the averages for cisgender men and cisgender women. This pattern is sometimes described as “sex-intermediate,” meaning the brain does not map cleanly onto either group but shows a blend of characteristics. The study’s significance lies partly in its size and partly in the fact that none of the transgender participants had started hormone treatment, ruling out the possibility that the differences were caused by exogenous hormones. Whatever was different in these brains was present before any medical intervention.

Brain imaging studies like this one should not be oversimplified into a narrative about having a “female brain in a male body” or vice versa. Brains are highly variable. There is enormous overlap between male and female brain structures in the general population, and brain sex is not a single dial that points in one direction. What the imaging work does show is that gender identity leaves a detectable footprint in brain anatomy, one that is present from the start rather than acquired through social experience or medical treatment.

What Natural Experiments Tell Us

Some of the most compelling evidence comes from people born with conditions that create a natural mismatch between their chromosomal sex and the hormones their bodies produce during development. Congenital adrenal hyperplasia (CAH) is one such condition. People with CAH who have XX chromosomes are exposed to unusually high levels of androgens in the womb, even though they are typically assigned female at birth.

A review of 250 adults with CAH who were assigned female found that about 5% experienced gender dysphoria and roughly 2% transitioned to living as men in adulthood.⁶ Those rates are far higher than in the general population of people assigned female. The conclusion is hard to miss: elevated prenatal androgen exposure shifted gender identity in some individuals despite being raised as girls from birth. It did not override identity in most cases, which again underscores that no single factor is deterministic, but the rate increase provides direct evidence that prenatal hormones influence gender identity in humans, not just in laboratory animals.

These natural experiments matter because they address a question that twin studies alone cannot fully answer. Twin studies show that something heritable is going on but leave room for debate about what exactly is inherited. The CAH data help fill in the mechanism: what is inherited, at least in part, is the sensitivity of the developing brain to sex hormones, and variations in that sensitivity can nudge gender identity in a direction that differs from what the rest of the body would predict.

Socialization Does Not Create Transgender Identity

A persistent misconception is that transgender identity is a product of how a child is raised, whether through parental influence, social media exposure, or peer dynamics. The evidence runs strongly against this. Transgender children who eventually socially transition were initially raised in a way that matched their sex assigned at birth. Photographs and family reports confirm that these children were dressed, named, and treated according to their assigned gender. Many families reported reacting negatively to their child’s gender nonconformity at first, not encouraging it.⁶

This matters because it directly tests the “social influence” hypothesis. If transgender identity were created by affirmation or by exposure to transgender role models, you would not expect it to emerge in children who were actively socialized in the opposite direction. Yet that is exactly what happens: children assert a gender identity that contradicts the one their families and communities tried to instill. The social environment may shape how comfortable a person feels expressing their identity, and it certainly affects whether they have the vocabulary and social support to articulate what they feel. But the evidence does not support the idea that socialization creates the identity itself.

How Stable Is Transgender Identity in Children?

One of the most contentious topics in public debate is whether children who identify as transgender will continue to do so as they grow up. Older studies, often cited in these discussions, suggested high rates of “desistance,” meaning children who appeared to outgrow their cross-gender identification. But those studies had serious methodological problems, including counting children who simply did not return for follow-up as having desisted, and using broad inclusion criteria that captured gender-nonconforming children who never actually identified as transgender.

More recent longitudinal research paints a different picture. A study that followed children who had made a full binary social transition found that approximately five years later, 94% still identified as transgender. About 3.5% identified as nonbinary, and only 2.5% had returned to identifying with the gender they were assigned at birth.⁶ For children whose gender identity is strong and consistent enough to prompt a social transition, that identity tends to persist. This does not mean every child who questions their gender will ultimately be transgender, but it does challenge the narrative that childhood transgender identity is typically a passing phase.

Why No Single Cause Has Been Found

People sometimes ask why, if gender identity is biological, scientists have not found “the gene” or “the brain region” that makes someone transgender. The answer is that gender identity, like most complex human traits, does not work that way. Height is about 80% heritable, but researchers have identified hundreds of genes that each contribute a tiny amount. Personality traits, cognitive abilities, and even preferences like being left-handed all involve many genes interacting with developmental environments. Gender identity fits this same pattern.

The Endocrine Society’s clinical practice guideline summarizes the state of the science by noting that evidence from genetics, endocrinology, and neuroanatomy all support the idea that gender identity reflects a complex interplay of factors, not a single biological switch.¹ The twin data show a genetic component. The prenatal hormone evidence shows a developmental component. The brain imaging data show that these influences leave traces in brain structure. And the natural experiment data from conditions like CAH help connect the dots between hormones and identity. No single study is a smoking gun, but together they build a consistent case.

The absence of a single neat cause is sometimes exploited rhetorically, as though complexity means uncertainty. But complexity is the norm for traits that involve the brain. The fact that multiple independent lines of evidence point in the same direction, genetics, hormones, brain anatomy, and clinical observations, actually makes the overall case stronger than it would be if it rested on a single finding. When different methods converge on the same conclusion, the conclusion is more trustworthy, not less.

What Current Research Still Lacks

The evidence base, while compelling in its direction, has genuine gaps. Most twin studies on gender identity rely on aggregated samples rather than population-level registries, which limits precision. Neuroimaging studies, including the ENIGMA analysis, are cross-sectional: they capture a snapshot of the brain at one point in time rather than tracking changes from before birth through childhood. And the molecular genetics research so far has focused on candidate genes, like the androgen receptor, rather than scanning the entire genome in large samples. Genome-wide association studies of the kind that have mapped the genetic architecture of height or depression have not yet been conducted at sufficient scale for gender identity.

There is also relatively little research on transgender men and nonbinary individuals compared to transgender women. Much of the early genetics and hormone research focused on people assigned male at birth, partly because the androgen receptor hypothesis lent itself to that direction. As the field matures, broader and more inclusive study designs are starting to fill in those blind spots. The hormonal theory predicts that the mechanisms should work in both directions, with lower-than-expected androgen exposure during brain development contributing to female gender identity in those assigned male, and higher-than-expected exposure contributing to male identity in those assigned female. The CAH data support the latter half of that prediction, but direct molecular evidence remains thinner for transgender men than for transgender women.

Epigenetics, the study of how gene expression is modified without changes to the DNA sequence itself, is another frontier. Prenatal stress, maternal immune responses, and placental function can all alter how genes are expressed during fetal development, and some researchers suspect these mechanisms could contribute to gender identity variation. The research here is early-stage and largely speculative, but it represents a plausible bridge between the genetic and hormonal findings, potentially explaining how identical twins with the same DNA can end up with different gender identities.

The Gap Between Scientific Evidence and Public Debate

Public conversations about why people are transgender often proceed as though the science is far more uncertain than it actually is, or as though biology and identity are in conflict. The research does not support either of those framings. The biological evidence is consistent across disciplines and has been building for decades. It is also compatible with the lived experience that transgender people routinely describe: a persistent, deeply felt sense of gender that does not match what was assigned at birth, present from early childhood and resistant to social pressure.

At the same time, the science does not reduce gender identity to a simple biological test, and no responsible researcher claims it does. There is no blood draw or brain scan that can diagnose someone as transgender. Self-report remains the gold standard for gender identity, just as it is for many other internal experiences that have biological underpinnings, such as sexual orientation, chronic pain, or even hunger. The fact that something relies on self-report does not make it less real or less biological. It just means the biology is operating at a level of complexity that current tools cannot fully capture from the outside.

A useful way to think about it: asking “why are people transgender” is a bit like asking “why are people left-handed.” There is a strong genetic component, a prenatal developmental component, and some environmental variability, and the result is a trait that is not chosen, not caused by upbringing, and not disordered. The analogy breaks down in some ways, handedness is simpler and less politically charged, but the underlying scientific logic is similar. Complex traits with biological roots do not always yield to simple explanations, and the absence of a simple explanation does not mean the trait is a mystery or a choice.