Category: Loss to Follow-Up
-
Three decades of research based on the Dutch Protocol have not produced reliable evidence to support the use of puberty blockers and hormones in adolescents.
In a 2026 critique, McDeavitt and Cohn reviewed three decades of research underpinning the Dutch Protocol and concluded that it has not produced reliable evidence to support the use of puberty blockers and cross-sex hormones in adolescents.
Their critique addressed a 2026 narrative review in which researchers from the Amsterdam Center of Expertise on Gender Dysphoria argued that three decades of their own research supported the use of puberty blockers and cross-sex hormones in adolescents. The review did not represent an independent evaluation of the evidence, as it was authored by the same research group that had generated much of the underlying literature.
The Amsterdam team concluded that puberty blockers and cross-sex hormones are associated with reduced gender dysphoria, high appearance and treatment satisfaction, improved mental health, good self-esteem and general functioning, low discontinuation rates of cross-sex hormones, and limited or no long-term adverse effects on growth following treatment with puberty blockers, hormones and surgery. They also reported that treatment from early puberty limits genital growth, meaning alternatives to penile inversion vaginoplasty may be required; that mastectomy at ages 16–17 produced similar satisfaction and low decisional regret compared with surgery at ages 18–20; and that adolescents were found able to make informed decisions about treatment. The Dutch authors concluded that the available evidence supports the provision of these interventions, while acknowledging that further research is needed in several areas.
McDeavitt and Cohn argued that these conclusions are not justified because the underlying studies have important methodological limitations. They identified inadequate long-term follow-up to assess outcomes such as cardiovascular disease, fertility and treatment regret; lack of appropriate comparison groups; reliance on surrogate outcome measures rather than clinically meaningful outcomes; substantial loss to follow-up (for example, in one study only 39% of participants could be contacted after seven years); inadequate adjustment for confounding variables; and the absence of any randomised controlled trials. They argued that these limitations prevent reliable conclusions about the long-term benefits and risks of treatment.
McDeavitt and Cohn concluded that, despite three decades of research, the evidence underpinning the Dutch Protocol remains insufficient to support the use of puberty blockers and cross-sex hormones in adolescents.
de Vries, A. L. C., Steensma, T. D., Kreukels, B. P. C., van de Grift, T. C., van Mello, N. M., & Hannema, S. E. (2026). The Dutch approach for transgender adolescents; past, present and future perspectives. European Journal of Developmental Psychology. https://doi.org/10.1080/17405629.2026.2615311
McDeavitt, K., & Cohn, J. (2026). Three decades of ‘Dutch Protocol’ research has not produced reliable evidence. European Journal of Developmental Psychology. https://doi.org/10.1080/17405629.2026.2680285
-
Complete national register follow-up eliminates loss-to-follow-up bias.
A nationwide Finnish study followed all individuals referred to gender identity services between 1996 and 2019 using mandatory national health registers, ensuring virtually complete follow-up with no attrition. By linking multiple administrative datasets over up to 25 years, it avoided common limitations of earlier research, including small samples, selective drop-out, and reliance on self-report. The study provides a comprehensive population-level assessment of long-term psychiatric treatment needs in this cohort.
Ruuska, S.-M., Tuisku, K., Holttinen, T., and Kaltiala, R. (2026). “Psychiatric Morbidity Among Adolescents and Young Adults Who Contacted Specialised Gender Identity Services in Finland in 1996–2019: A Register Study,” Acta Paediatrica: 1–9. https://doi.org/10.1111/apa.70533.