Michael Bailey

Finding neurological differences in transgender people does NOT establish that there is some innate “gender identity” that we all possess. Look at depression. We could hypothetically identify specific changes in neural circuitry associated with depression, but this doesn’t mean we’ve identified some dedicated “depression circuit” that everyone possesses. Instead, these differences may reflect alterations in systems involved in a wide range of other functions. Why should we interpret neurological differences in transgender populations any differently? Even if we identified consistent neurological differences, these could reflect variation in neural systems involved in numerous other functions, rather than evidence of a dedicated “gender identity” mechanism that everyone possesses. Finding neural correlates of a phenomenon does not establish that it represents a distinct, universal psychological characteristic.

One thing that genuinely amazes me after working in Alzheimer’s research is seeing people describe the transgender “brain-sex” literature as particularly strong science. Look at what high-quality biomarker research in AD actually looks like. Cohorts of hundreds or thousands of participants, longitudinal follow-up, neuropathological or PET/CSF validation, independent replication across cohorts and laboratories, head-to-head comparisons between biomarkers, careful adjustment for confounders, and increasingly impressive analytical validation. It has taken YEARS of work like this before researchers have become comfortable saying that biomarkers such as plasma p-tau217 can reliably reflect underlying AD pathology. Then compare that standard with some of the studies routinely cited as showing that transgender people have brains shifted towards the opposite sex. Some of the classic post-mortem studies are based on literally single-digit numbers of transgender brains. Participants are frequently exposed to long-term cross-sex hormones. Sexual orientation is often inadequately controlled despite itself being associated with some of the same neuroanatomical measures. Some findings don’t replicate across neighbouring hypothalamic nuclei, and even supposedly sexually dimorphic regions are not always consistently sexually dimorphic across studies. The neuroimaging literature isn’t exactly reassuring either: generally small, heterogeneous samples, different analysis pipelines and regions of interest, hormone-naïve and hormone-treated participants mixed across the literature, and results that can variously look sex-typical, intermediate, “cross-sex shifted”, or simply atypical depending on what measure you examine. Yet from this incredibly messy literature comes the remarkably confident claim that transgender people have a “female brain in a male body” (or vice versa), or that neuroimaging has demonstrated a biological brain basis for “gender identity”. The difference in standards is genuinely striking. In AD research, people have spent enormous amounts of time, money and effort establishing whether a blood test can accurately tell us something as comparatively concrete as whether amyloid or tau pathology is present in the brain. Apparently demonstrating that an enormously complex psychological construct originates from prenatal sexual differentiation of the human brain requires considerably less…