Taux de détransition inconnus : Une revue systématique expose des lacunes critiques dans les données à long terme

Prévalence de la détransition chez les personnes recherchant des traitements hormonaux d'affirmation de genre : une revue systématique

Revues systématiques & des données probantes Détransition & regret

Authors: Eva Feigerlova
Journal: Journal of Sexual Medicine
Year: 2025
Source: pubmed.ncbi.nlm.nih.gov/39724926/

Summary

This review looked for studies measuring how often people who had asked for or started puberty blockers or cross-sex hormones later reversed course. It included 15 clinic, registry and insurance studies covering about 7,000 people. Most reported low figures, usually under 10%. But nearly all studies looked back at old records, used different meanings of detransition, and followed people for a short or unknown time. They also did not separate a change in identity from stopping because of cost, side effects or other reasons. One US military health study found about 30% stopped hormones within 4 years, a figure the review's headline range leaves out. The author concludes detransition has not been studied enough and calls for long-term studies that follow people forward. The ranges are best read as minimum figures from short follow-up, not reliable rates.

Key findings

La revue incluait 15 études portant sur 3 804 enfants et adolescents et 3 270 adultes ; 14 examinaient d'anciens dossiers et une suivait les personnes dans le temps sans groupe de comparaison.
Avant tout traitement, 0,8 % à 7,4 % ont changé leur demande ; 1 % à 7,6 % ont arrêté les bloqueurs de puberté ; et 1,6 % à 9,8 % ont arrêté les hormones.
Une étude de santé de l'armée américaine (Roberts 2022) a trouvé qu'en 4 ans, 19 % des femmes trans et 35,6 % des hommes trans ont arrêté les hormones, bien au-dessus de la fourchette principale de la revue.
Les définitions variaient largement, de ne plus s'identifier comme trans au simple fait de rester plus de 90 jours ou 13 mois sans hormones.
Les raisons d'arrêt comprenaient des doutes sur l'identité, mais aussi des effets secondaires, le coût, des problèmes sociaux, l'atteinte des objectifs du traitement et l'absence de suivi ; dans une étude d'une clinique britannique, la raison était inconnue pour 10 des 12 personnes qui avaient arrêté.
Globalement, les preuves ont été jugées de très faible qualité, et la plupart des études suivaient les personnes trop brièvement ou ne rapportaient pas du tout le suivi.
La revue a écarté les enquêtes en ligne auprès de détransitionneurs, parce que les personnes qui choisissent d'y répondre peuvent ne pas être typiques.
L'auteur conclut que la détransition n'a pas été assez étudiée et appelle à des définitions communes et à des études à long terme qui suivent les personnes dans le temps.

Abstract

Background: Despite recent evidence of the benefits of gender-affirming medical procedures, data in the literature indicate emerging demands of detransition and regrets while suggesting potential sources of bias in different datasets, including a nonconsensual definition of detransition. Aim: The present systematic review aims to summarize the existing research regarding the prevalence of detransition in transgender persons who requested or started receiving gonadotrophin-releasing hormone analogs (GnRHa) and/or gender-affirming hormonal therapy (GAHT). Methods: A systematic literature search (CINAHL Plus, Cochrane Library, Google Scholar, MEDLINE, Web of Science, Sage Journals, Science Direct, Scopus) for quantitative studies was conducted up to May 2024. All eligible studies were assessed using the Strengthening the Reporting of Observational Studies in Epidemiology statement. The risk of bias was assessed using the National Institute of Health quality assessment tool. The present analysis follows the PRISMA statement for systematic review articles and the synthesis without meta-analysis recommendations. Outcomes: The primary outcome was the point-prevalence proportion of detransition events as a percentage in the population of transgender persons who were considered eligible for treatment or had initiated GnRHa and/or GAHT. Results: Fifteen observational studies involving 3804 children and adolescents and 3270 adult participants were included in the 3212 screened studies. Five studies reported a change in request before starting GnRHa, five studies reported GnRHa discontinuation, and nine studies reported GAHT discontinuation. The point-prevalence proportions of shifts in requests before any treatment ranged from 0.8-7.4%. The point-prevalence proportions of GnRHa discontinuation ranged from 1-7.6%. The point-prevalence proportions of GAHT discontinuation ranged from 1.6-9.8%. All of the included studies were heterogeneous regarding definitions of detransition used and the study design: their numbers were too small to be statistically relevant, their time frame was insufficient, they did not use patient-level data, or they did not consider confounding factors. Clinical implications: Quality measurement tools are needed, as are monitoring standards, and both are important for health outcomes and guarantee the attention of health care providers and policy-makers. Strengths and limitations: The unique features of this analysis are its restrictive inclusion criteria compared with those of previous reviews, such as a strict definition of detransition and a focus on empirical studies only. However, most of the studies were retrospective and unblinded, and most were not sufficiently powered to detect detransition rates. Conclusion: Taken together, the results of the present analysis show that detransition in persons undergoing gender-affirming treatment has been insufficiently investigated, highlighting the need for long-term follow-up studies.

Conclusion

This review provides the point-prevalence proportions of detransition among persons who sought hormonal treatments (GnRHa and/or GAHT). Taken together, the results of the present analysis indicate that detransition in persons undergoing gender-affirming treatment has been insufficiently investigated. Quality measurement tools are needed, as are long-term follow-up and monitoring standards. The present study highlights the need for well-designed prospective studies to evaluate the number of transgender persons experiencing detransition after starting GnRHa and/or GAHT and associated factors and to detect individuals at a greater risk of detransition. These data are important for health outcomes and to guarantee the attention of health care providers and policy makers.

Methodological limitations

Only one reviewer: One researcher found, read and rated all the studies. The author admits this raises the risk of mistakes and bias compared with the usual two reviewers.
Headline range leaves out the highest figure: The 1.6% to 9.8% hormone range does not reflect Roberts et al. (2022), which is in the review. It found about 30% (19% of trans women, 35.6% of trans men) stopped hormones within 4 years in the US military health system. A gap in prescriptions is not always detransition, but the same is true of other studies in the review.
Very weak studies: 14 of 15 studies looked back at old records, mostly without any blinding, and the overall evidence was rated very low quality.
Short or unknown follow-up: Most studies followed people only briefly, but detransition often happens years later (in one Finnish clinic study, regret came after 7 years on average). So the figures are likely too low.
Different definitions: Detransition ranged from no longer identifying as trans to a 90-day gap in prescriptions. So the ranges lump together different things.
Clinics miss many cases: Studies only count people known to their clinic or insurer. Surveys suggest many detransitioners never tell their original clinic, and the review does not adjust for this.
Reasons often unknown: Several studies did not record why people stopped (for example, 10 of 12 in Hall 2021). So stopping because of identity cannot be separated from stopping for other reasons.
Second-hand data: Five studies used existing datasets or reports rather than individual patient records, and in seven it was unclear exactly what was measured.
Surveys left out: Online detransitioner surveys were excluded as biased. That is reasonable, but it leaves out the main source of information on why people who left clinic care stopped, and what they went through.
Very different groups combined: Children, teens and adults, and different countries' treatment rules, are mixed together. Sex and age could not be worked out for many studies.
Framing goes beyond the data: The review opens by citing 'recent evidence of the benefits' and says detransition 'appears to be rare'. These claims are stronger than its own very weak data can support.

detrans.ai

Summary

Key findings

Abstract

Conclusion