What this review does not conclude

Evidence Review · ER-015 · Reader’s note

This page exists because the review is about to be misquoted, and it will be misquoted in both directions.

Masculinising and Feminising Hormones in Adolescents (ER-015) grades the evidence on testosterone and oestradiol in adolescents with gender dysphoria or incongruence. It answers an evidence question. It deliberately does not answer the policy question, and it says so at length.

What follows is not a summary. It is a list of things the review does not say — set out here so that anyone can check a quotation against it in thirty seconds.

It does not conclude that these medicines do not work

Very low certainty is a statement about the studies, not about the drugs. All three formal appraisals — NICE in 2020, the systematic review commissioned for the Cass Review in 2024, and NHS England’s evidence reviews in 2026 — rated the evidence very low certainty or lacking in high-quality studies. None of them concluded that the treatment does not work. Neither does this review.

NICE’s own conclusion, at very low certainty, was that hormones are “likely to improve symptoms of gender dysphoria.” That finding is weak. It is not absent.

It does not conclude that these medicines are safe

Absence of evidence is not evidence of absence — in either direction. Fertility in transmasculine adolescents: unstudied. Cognitive development: effectively unstudied. Lifetime regret: unstudied. Cardiovascular events: unstudied. Cancer: the latency is decades and the cohorts are years old.

An evidence base that has not looked cannot reassure. The review says so about every one of those outcomes.

It does not recommend withdrawing the treatment

“The evidence is weak” is an evidence claim. “Therefore the treatment should be withdrawn” is a policy claim, and the first does not entail the second. Weak evidence can justify caution, more research, restriction, research-only provision, or continued provision with informed consent — depending on values the evidence cannot supply.

It does not recommend unrestricted access

“Young people report relief” is also an evidence claim, of limited certainty, and “therefore access should be unrestricted” is a policy claim it cannot carry. The evidence for benefit is observational, uncontrolled, and cannot establish causation — and no agreed threshold exists for deciding whether an observed improvement is large enough to matter.

It does not take a position on the NHS England consultation

The review was written to stand whichever way that decision lands. If it reads, in retrospect, as having been written to anticipate one outcome, it has failed.

It does not treat the errors it found as a scandal

Sixteen apparent errors in NHS England’s evidence reviews were reported to the NHS England Clinical Effectiveness Team on 12 July 2026, before this review was published. Correcting them does not change the certainty ratings — and several of them, corrected, would strengthen rather than weaken the case for caution.

The review’s own correction record runs to thirteen entries, published in full, including errors that cut against the author’s stated position. The direction of an error is not the reason it is recorded. How that record is kept, and where it failed, is set out in the Deposit Standard.

What it does conclude

The physical effects are firm. Testosterone virilises; oestradiol feminises. Several of those changes substantially persist after treatment stops — and so do several of the changes produced by a puberty allowed to continue. Neither treating nor deferring is free of lasting consequence.
The psychological evidence is observational, uncontrolled and of very low certainty. No randomised trial exists; no controlled study capable of establishing causation was identified by any of the three appraisals.
The treatment’s central intended outcome has been measured once, in twenty-three people — and that study administered one version of the measuring instrument before treatment and a different version afterwards. Its headline figure cannot be read as a conventional before-and-after change. The paper disclosed this. No subsequent evidence summary carried it forward.
Pelvic pain, documented in roughly one in four transmasculine adolescents on testosterone, is not a named outcome in any of the ten commissioned reviews.
No review contains any outcome for what happens when treatment stops — although the same documents state that these medicines mitigate the effects of hypogonadism, and name its symptoms.

The two sentences that will be quoted against each other.

“The evidence is weak” is true, and it does not entail that the treatment should be withdrawn.

“Young people report relief” is also true, and it does not entail that access should be unrestricted.

They are both in this review, at the same size, on purpose.

How to cite, and how to check

Eden. Masculinising and Feminising Hormones in Adolescents — Evidence Review (ER-015). Eden Openly, 2026. Version 1.3. doi:10.5281/zenodo.21329685

Cite the review, not this page. This is a reader’s note. Everything on it is drawn from the review, and every claim in the review is graded and referenced.

The full review, its declarations, its provenance statement and its complete correction record are on the canonical page. A short companion Evidence Note distils it in eight minutes.

If you find an error, tell us. It will go in the correction record, with your finding credited, whichever way it cuts.