Key Findings

• Concludes the analyses do not adequately support Kurth et al. (2022)'s headline claim that transgender women's brains are shifted away from their biological sex toward their gender identity
• Only 24 participants per group: effect sizes like d = 0.64 are unstable at n < 30, and an identified outlier in the cisgender male group was never sensitivity-tested
• Lilliefors normality tests used to justify parametric statistics have under 14.5% power at n ≤ 30, so passing them does not establish normality — nonparametric Kruskal–Wallis tests were called for
• One-tailed post hoc t-tests halved the key p-value (0.032 to 0.016); Bonferroni correction across the three comparisons raises it to 0.048, and the conventional two-tailed corrected value reaches 0.096, erasing significance for the principal transgender-women vs cisgender-men comparison
• Sexual orientation never analyzed as a neuroanatomical confound despite the cohort being 6 androphilic and 18 gynephilic and the authors possessing the orientation data — prior literature shows orientation associates with brain structure
• Given the paper's unusual reach into media, law, and policy, the authors call for independent replication and public release of the anonymized participant-level data before it is cited as evidence

Abstract

Methodological peer review finding that Kurth et al. (2022) brain-sex claims rest on flawed statistics and overstated conclusions.

Summary

A 2026 peer review found that Kurth et al. (2022), widely cited for the claim that transgender women's brains shift toward their gender identity, relied on flawed statistical methods and ignored sexual orientation as a confound, overstating its support for the hypothesis.

Conclusion

The evidence presented does not justify the headline conclusion that brain anatomy in transgender people is shifted away from their biological sex toward their gender identity.