Home › Science & debate › Research and statistics

Research and statistics

This overview brings together the main scientific studies and statistics on gender dysphoria and medical gender care. The common thread is that the evidence base is considerably weaker than the certainty of many publication texts and care guidelines suggests. Systematic reviews from the UK, Sweden and Finland conclude independently of each other that the quality of evidence for the central interventions — puberty blockers and cross-sex hormones in minors — is predominantly very low.

Prevalence and the explosion of referrals

Clinical studies from the 1960s–1990s estimated the prevalence of gender dysphoria at 1 in 10,000 to 1 in 30,000, with a clear preponderance of biological males. Recent self-report studies — with much broader definitions — arrive at 0.5 to 1.7% of the population. More important than this absolute figure is the demographic shift: since around 2010 the number of referrals to youth gender clinics in the Netherlands, the UK and Scandinavia has multiplied tenfold to twentyfold. The sex ratio has flipped: the majority of current presenters are adolescent girls, often with co-morbid autism, anxiety or depression. No existing biological explanation can plausibly account for such a rapid, geographically bounded and sex-specific rise; socio-cultural factors, social media and peer contagion are increasingly cited as the main explanation.

Effectiveness: what does the evidence really say?

The Cass Review (2024) had the University of York carry out systematic reviews according to GRADE criteria. Conclusion: the quality of evidence for puberty blockers and cross-sex hormones in young people was predominantly "very low". No convincing evidence was found that blockers improve psychological well-being, lower suicidality or "buy time". Earlier systematic reviews by the Swedish SBU (2022) and the Finnish COHERE (2020) independently reached the same conclusion. Levine, Abbruzzese and Mason (2022) document methodological problems in WPATH SOC-8, including the removal of minimum age limits without an evidence base.

The Dutch Protocol re-examined

Amsterdam UMC published between 2006 and 2014 the original outcome studies that made the Dutch Protocol known worldwide. The sample was small (n=55 for the main publication), without a control group, with high loss to follow-up, and with cherry-picking of outcome measures. One participant died during vaginoplasty; this was not mentioned in the main publication. A British attempt at replication (Carmichael et al., Early Intervention Study) found no improvement on psychological outcome measures. Generalisation of the protocol to the current patient population — adolescent girls with co-morbid problems rather than young children with early persistent dysphoria — is not scientifically justified.

Long-term outcomes and mortality

A Swedish cohort study by Dhejne et al. (2011) found significantly increased mortality and suicide risks in adult post-operative trans people compared with the general population, even ten years after transition. The authors emphasised that this underlines the need for continued psychiatric care, but it also refutes the idea that transition systematically "cures" psychological complaints. A large Danish register analysis (2024) again found excess mortality. The absence of RCT evidence and the short follow-up of most studies make strong causal claims problematic.

Detransition: likely far higher than reported

For decades detransition figures of <1% were published. These figures rest on clinical cohorts with high loss to follow-up and short observation durations. Recent self-report studies and cohorts with longer follow-up report percentages between 7 and 30%. Detransitioners often state that underlying problems — trauma, internalised homophobia, autism, dissociation — were not adequately investigated before medical treatment. See detransition.

Statistics in the Netherlands

Amsterdam UMC reported a tenfold increase in referrals among young people between 2010 and 2020. Waiting lists have been growing for years. CBS (Statistics Netherlands) registers legal sex changes, but clinical outcome data are not centrally and transparently published. As a result, an independent long-term evaluation of Dutch practice is effectively impossible — a scientific and governance gap that has been noted by, among others, the National Health Care Institute (Zorginstituut Nederland).

Methodological problems in summary

The common thread in the criticism: small and highly selected samples, no control groups, short follow-up, high loss to follow-up, changing and favourably selected outcome measures, and publication bias. In addition: a research field in which dissent was until recently discouraged and in which patient organisations and activist advocates weigh heavily in guideline committees. Good medicine requires that uncomfortable findings are published as much as confirmatory ones.