On 12 February 2026 the U.S. Food and Drug Administration approved labeling changes that removed cardiovascular disease, breast cancer, and probable dementia from the boxed warning on six menopause hormone therapy products.1 No trial reported new results that week. A drug label is a communication decision made about existing evidence, and understanding what actually changed — and what was deliberately left in place — is the difference between an informed decision and a marketing message.
Key facts
- The FDA removed cardiovascular disease, breast cancer, and probable dementia from the boxed warning, and removed the instruction to use the lowest effective dose for the shortest possible time.12
- The cardiovascular and breast cancer warnings were retained elsewhere in the label. They moved out of the box; they were not deleted.2
- A boxed warning for endometrial cancer remains on systemic estrogen-alone products.2
- Hormone therapy is the most effective treatment available for hot flashes and night sweats, and that was true before the label changed.3
- Over 18 years of follow-up, the Women’s Health Initiative found no increase in all-cause mortality among women who had taken hormone therapy.6
- Age at initiation changes the risk arithmetic more than any other single variable.7
Start here
Hormone therapy evidence
Benefits, risks, WHI without the mythology, route and timing, and the 2026 FDA label context.
Perimenopause vs menopause
Definitions, what actually changes, and when testing helps versus confuses.
Testosterone in menopause
A separate question with a much narrower evidence base and no FDA-approved product for women.
GLP-1 drugs and menopause
Why midlife weight medications and HRT are different evidence files — not one lifestyle stack.
What the FDA changed, and what it kept
The agency began this process in November 2025, after a July 2025 expert panel, a public comment period, and what it described as a comprehensive assessment of the literature published since the Women’s Health Initiative.2 Twenty-nine drug companies submitted proposed labeling changes; the first six approvals span all four categories of menopause hormone therapy.1
Coverage of the decision has been lopsided. The removals were widely reported; the retentions much less so. Both are in the same FDA document.
| Removed | Kept or added |
|---|---|
| Cardiovascular disease, breast cancer, and probable dementia language in the boxed warning, for all products | Cardiovascular disease and breast cancer warnings retained in the labeling as a whole |
| Endometrial cancer language in the boxed warning — except for systemic estrogen-alone drugs | Boxed warning for endometrial cancer retained on systemic estrogen-alone products |
| The recommendation to use the lowest effective dose for the shortest amount of time | Added: consideration of starting therapy for moderate to severe hot flashes in women under 60 or within 10 years of menopause |
| The probable dementia warning throughout the label | Added: Women’s Health Initiative data specific to women aged 50–59 |
Two things are worth holding at once. The FDA followed a defined regulatory process, and the direction of the change is broadly consistent with how specialist bodies have read the evidence for over a decade.3 It is also true that no new randomized trial prompted it, and that the announcement was delivered at a Department of Health and Human Services press conference framed around correcting past messaging.1 A label change is a regulatory judgment about how to describe known evidence. It is not itself evidence.
What hormone therapy is actually approved to treat
Menopause hormone therapy has three well-established uses: moderate to severe vasomotor symptoms — hot flashes and night sweats — genitourinary symptoms caused by falling estrogen, and prevention of bone loss.13 For hot flashes specifically, nothing else performs as well.
It is not approved, and the evidence does not support prescribing it, as a general anti-ageing intervention, a treatment for fatigue in isolation, or a way to prevent chronic disease in women who have no symptoms. That distinction matters because the same product can be entirely appropriate for one woman and unsupported for another with a superficially similar complaint.
The Women’s Health Initiative, misread twice
In 2002 the estrogen-plus-progestin arm of the Women’s Health Initiative was stopped early after investigators reported an increased risk of breast cancer; the estrogen-alone arm was stopped in 2004 after an increase in stroke and no evidence of coronary protection.4 Prescribing collapsed. Those findings were real for the population studied.
The population is the point. Participants averaged 63 years old, and many were more than a decade past menopause — a group in which the balance of benefit and risk differs substantially from a symptomatic 52-year-old. Longer follow-up sharpened the picture considerably. Across the intervention and extended post-stopping phases, risks and benefits varied by outcome and by age.5 After 18 years of cumulative follow-up, hormone therapy was not associated with an increase in all-cause mortality, or with mortality from cardiovascular disease or cancer.6
So the trial was misread twice. First when its results were generalized to every woman at every age, and now, in some quarters, when the correction is presented as though the original safety signals never existed at all. The honest summary is narrower and duller than either version: the risks are real, they are smaller in absolute terms than the 2002 coverage implied, and they depend heavily on who is being treated.
Timing changes the arithmetic
An analysis of cardiovascular risk by age and by years since menopause found that the relationship between hormone therapy and cardiovascular disease differs according to how long a woman has been postmenopausal.7 This is the observation underlying what clinicians call the timing hypothesis, and it is why current guidance frames the decision around age and proximity to menopause rather than treating hormone therapy as uniformly safe or uniformly risky.3 The FDA’s 2026 labeling explicitly adds this framing, directing attention to women under 60 or within 10 years of menopause.2
It also means a claim like “hormone therapy is safe” is incomplete to the point of being useless. Safe for whom, started when, by which route, and weighed against which symptoms.
Where the evidence is strong, and where it is not
| Claim | What the evidence supports |
|---|---|
| Relieves moderate to severe hot flashes and night sweats | Strongly supported; the most effective available treatment3 |
| Relieves genitourinary symptoms of menopause | Strongly supported, including low-dose vaginal preparations11 |
| Prevents bone loss and reduces fractures | Supported by randomized data15 |
| Does not increase all-cause mortality over long follow-up | Supported by 18-year follow-up of a large randomized trial6 |
| Oral preparations carry a higher clotting risk than transdermal | Supported by large observational data, not by randomized trials8 |
| Should be started to prevent heart disease in women without symptoms | Not supported; this was the original WHI question and it was not met4 |
| Carries no breast cancer consideration now that the box has changed | Not supported; the warning was retained in the label2 |
Route is not a technicality
A large nested case-control study using two UK primary care databases found that oral hormone therapy was associated with an increased risk of venous thromboembolism, while transdermal preparations were not associated with increased risk.8 This is observational evidence, so it cannot establish causation with the confidence a randomized trial would, and we grade it accordingly. It is nonetheless consistent and clinically influential, and it is the reason route of administration is a substantive part of the conversation rather than a matter of preference.
If you would rather not take hormones
Hormone therapy is contraindicated for some women and unwanted by others, and the non-hormonal evidence base is no longer thin. Current guidance supports several options, including cognitive behavioural therapy and clinical hypnosis alongside specific medications.9 Among newer drugs, fezolinetant — a neurokinin-3 receptor antagonist that works on the brain’s temperature regulation rather than on estrogen receptors — reduced the frequency and severity of moderate to severe hot flashes compared with placebo in a phase 3 randomized trial.10
Genitourinary symptoms deserve separate mention because they behave differently from hot flashes: they tend to persist or worsen over time rather than resolve, and low-dose vaginal estrogen delivers minimal systemic exposure.11 A woman who declines systemic therapy has not necessarily declined this.
Questions worth asking
- Which symptoms are we actually treating, and how will we know in three months whether it worked?
- How many years past my final period am I, and how does that change the calculation for me specifically?
- Systemic or local — and if systemic, oral or transdermal, and why that route for me?
- Do I have a uterus, and if so what is protecting the endometrium?
- What in my personal or family history — clotting, stroke, breast cancer, liver disease — changes this decision?
- What is the plan for reviewing this, and what would make us stop?
Frequently asked questions
Did the FDA decide hormone therapy is safe?
No. It decided that three specific risks no longer belonged in the most prominent warning box. The cardiovascular and breast cancer warnings remain in the label, and a boxed warning for endometrial cancer remains on systemic estrogen-alone products. The change was to how risk is communicated, not a finding that risk is absent.
Was there new research behind the 2026 change?
No new trial was published to prompt it. The FDA reviewed literature accumulated since the Women’s Health Initiative, held an expert panel in July 2025, and sought public comment. The underlying evidence had been available and debated for years; what changed was the agency’s judgment about how to summarise it on the label.
Is it too late for me to start?
That depends on your age and how long it has been since your final period, which is exactly the variable current labeling and guidance emphasise. The benefit-risk balance is most favourable for women under 60 or within 10 years of menopause. Beyond that window it is not automatically ruled out, but it is a different and more individual conversation.
Does hormone therapy cause breast cancer?
The estrogen-plus-progestin arm of the WHI reported an increased risk of breast cancer diagnosis, and that warning remains in the label. The absolute size of the increase, how it varies by preparation and duration, and how it compares with other modifiable risks are all part of a proper discussion. What is not accurate is either claiming the risk is settled and large, or claiming the 2026 label change eliminated it.
Do I need a blood test to diagnose menopause?
Usually not. In women over 45 with typical symptoms and changing cycles, menopause is a clinical diagnosis, and hormone levels fluctuate enough during the transition that a single measurement can mislead. Testing has specific uses, particularly when menopause occurs early or the picture is unclear.
How long can someone stay on it?
The FDA removed the blanket instruction to use the lowest dose for the shortest time, which had been widely read as a hard stop. That does not convert into an open-ended prescription. Duration is now framed as an individual decision to be reviewed periodically rather than a fixed limit applied to everyone.
The bottom line
Menopause hormone therapy is an effective treatment for specific symptoms, with real risks that vary substantially depending on a woman’s age, her time since menopause, the route she takes it by, and her own medical history. That statement was accurate in 2024 and it is accurate now. What changed in February 2026 is that the label communicates it differently — less alarmingly, and in several respects more accurately.
Be careful with anyone using the FDA’s decision as a reason you should start. The agency corrected how a set of risks is presented. It did not discover that they were never there, and it left the cardiovascular and breast cancer warnings standing in the label precisely because they still apply.