What the studies show
Cohort and registry studies supply most of this field's hard numbers on diagnoses, treatment continuation, side effects and deaths. Population records show a steep rise in diagnoses, about 50-fold in English GP records from 2011 to 2021 (Jarvis et al., 2025). National mortality studies agree that death and suicide rates are raised. All-cause mortality was roughly double that of matched controls in Sweden (Clark et al., 2026), and suicide was 3.5 times higher in Denmark (Erlangsen et al., 2023). Where psychiatric history was examined, it explained much of the gap. In Finland, the excess suicide risk became compatible with chance once specialist psychiatric care was accounted for (Ruuska et al., 2024). In Taiwan, adjusting for depression shrank the suicide link the most (Hsu et al., 2025). Clinic cohorts also report physical risks, such as 4.5 to 5.5 times the expected rate of venous blood clots in trans women (Nota et al., 2019).
Uncontrolled cohorts of youth on puberty blockers found no measurable change in gender distress (de Vries et al., 2011; Carmichael et al., 2021) or in depression over two years (Olson-Kennedy et al., 2025), and nearly all went on to cross-sex hormones. Continuation figures conflict. In Amsterdam, 98% of youth still had a hormone prescription at around age 20 (van der Loos et al., 2022), while about 30% stopped within four years in US military records (Roberts et al., 2022).
Limits of the evidence
Most of these studies lack an untreated comparison group. Registry studies typically match only on age and sex, so they show that risks are raised but not why. Key estimates often rest on a handful of events, such as 12 suicides in Denmark. Regret is measured through narrow proxies over short follow-up. In Sweden, 21 of 2,467 people reversed a legal gender change (Clark et al., 2025). In Amsterdam, regret took about 11 years on average to appear, and 36% of patients stopped attending the clinic (Wiepjes et al., 2018). Many cohorts are clinics reporting on their own care. One Norwegian cohort was retracted on ethics grounds (Nyquist et al., 2025), and a widely cited surgery study was corrected after a reanalysis found no advantage of surgery (Bränström & Pachankis, 2020).