Hormone therapy: what the guidelines say, and where they stop

Few medical topics have been handled worse in public than hormone therapy. For two decades the message was a blanket warning. Lately, in some quarters, it has swung to something close to advocacy. Neither reflects the guidelines, which are more conditional and more useful.

This is not treatment advice. It is an account of what two major guidelines state, so you can recognise the reasoning when a clinician uses it and ask better questions when they do not. NICE NG23 covers identification and management of menopause, including premature ovarian insufficiency ⁶. Non-hormonal options for hot flushes are covered separately on this site, drawing on the 2023 nonhormone position statement ⁸.

What HRT is for

Both guidelines are consistent on the primary indication.

NICE recommends offering HRT to people with vasomotor symptoms associated with menopause ¹. The Menopause Society states that hormone therapy remains the most effective treatment for vasomotor symptoms and for the genitourinary syndrome of menopause, and that it has been shown to prevent bone loss and fracture ².

Both are equally consistent on what it is not for. NICE says do not offer combined or oestrogen-only HRT for primary or secondary prevention of cardiovascular disease ¹, and do not offer it for the purpose of dementia prevention ¹.

The timing argument

The single most important variable in the modern guidelines is when you start.

The Menopause Society position is that for women aged under 60 or within ten years of menopause onset who have no contraindications, the benefit-risk ratio is favourable for treating bothersome vasomotor symptoms and preventing bone loss. For women who start hormone therapy more than ten years from menopause onset, or who are older than 60, the benefit-risk ratio appears less favourable, because of the greater absolute risks of coronary heart disease, stroke, venous thromboembolism and dementia ².

Note the phrase "greater absolute risks". These are risks that rise with age regardless of hormone therapy. That is a large part of why the same drug looks different at 52 and at 68.

NICE reaches a compatible position by a different route, presenting outcome-by-outcome tables and requiring clinicians to share them, and noting that dementia risk might increase with combined HRT if it is started at 65 or over ¹.

Type matters: combined versus oestrogen-only

If you have a uterus, you need a progestogen alongside oestrogen. This is not a formality.

NICE says offer combined HRT to people with a uterus, and oestrogen-only HRT to people who have had a total hysterectomy ¹. The reason is in its outcome table: in people with a uterus, endometrial cancer risk increases with oestrogen-only HRT, with both oral and transdermal routes ¹.

The Menopause Society describes the same mechanism from the other side. When adequate progestogen is combined with systemic oestrogen, the risk of endometrial neoplasia is not higher than in untreated women ³.

There is a counterintuitive corollary. Continuous combined HRT decreases endometrial cancer risk relative to no HRT, while sequential combined HRT may slightly increase it, with the increase potentially greater with longer duration, fewer days of progestogen per cycle, and higher oestrogen dosage ¹.

The NHS notes that progestogen can be taken as tablets, as part of a combined patch, or delivered through a hormonal coil ⁵⁷.

Route matters: transdermal versus oral

This is the practical distinction most worth understanding, because it is where the guidelines diverge in confidence.

NICE's tables state that venous thromboembolism risk is not increased with transdermal HRT, that VTE risk is increased with oral HRT, and that VTE risk is greater with oral than transdermal ¹. On stroke, it states that stroke risk is unlikely to increase with transdermal oestrogen, while stroke risk increases with oral oestrogen, rising with higher dose, longer duration, and greater age at first starting, and differing between ethnic groups with a possibly greater increase in Black people ¹.

NICE also gives a direct clinical instruction: consider transdermal rather than oral HRT for people at increased risk of venous thromboembolism, including those with a body mass index over 30 ¹.

The Menopause Society is more cautious about the underlying evidence. Its 2022 statement notes that non-oral routes may offer potential advantages because they bypass first-pass hepatic metabolism, but says it is unknown whether non-oral routes are associated with lower risk of VTE, breast cancer and cardiovascular events compared with oral routes, because the clinical trials have not been done ³. It also notes that trials directly comparing myocardial infarction, stroke, breast cancer and VTE risk across estrogen routes and doses are lacking ³.

Both positions are defensible. NICE is drawing on the available observational evidence to make a recommendation clinicians can act on. The Menopause Society is pointing out that the randomised evidence is absent. If your clinician offers a patch rather than a tablet because of your BMI or clotting history, that is NICE's reasoning. If another clinician says the route difference is not proven, they are citing the same gap the Menopause Society flags.

For relieving hot flushes, meta-analysis found no significant difference between transdermal and oral combined therapy ³.

Breast cancer, stated carefully

This is the risk that dominates the conversation, so it is worth quoting the guidelines closely rather than paraphrasing.

NICE, on combined HRT: breast cancer risk increases with combined HRT, and the increase rises with duration of use, is higher in current than past users, and declines after stopping but persists at least ten years after stopping. There is a very small increase in risk of death from breast cancer with combined HRT. Risk with sequential combined HRT is lower than with continuous combined HRT, but higher than without HRT ¹.

NICE, on oestrogen-only HRT: there is very little or no increase in breast cancer risk, and little or no increase in breast cancer mortality ¹.

NICE also states that breast cancer risk varies depending on a person's own modifiable and non-modifiable risk factors, and that there is insufficient evidence to establish whether the increase in risk differs with preparations containing micronised progesterone or dydrogesterone compared with other progestogens ¹. That last point is worth knowing, because claims about "body-identical" progesterone being risk-free on this outcome go beyond what NICE says is established.

The Menopause Society's account of the trial data is compatible and adds nuance. In the active phase of the Women's Health Initiative, breast cancer incidence was higher in women assigned to conjugated equine estrogens plus medroxyprogesterone acetate compared with placebo, but reduced in women assigned to estrogen alone compared with placebo, and after a median 20 years of follow-up both of those directions persisted ³. Observational data, in contrast, have shown increased risk with either systemic regimen, in a duration-dependent way ³.

Putting the numbers in proportion

The Menopause Society makes a point about how risk should be communicated: absolute risks are more useful than relative risks in clinical conversations ³.

Its own summary is that the increased absolute risks associated with combined and oestrogen-only therapy are rare, defined as fewer than 10 events per 10,000 women per year, and include increased risk of VTE and gallbladder disease, with combined therapy carrying a rare increased risk of stroke and breast cancer ³. It also states that absolute risks are reduced for all-cause mortality, fracture, diabetes, and for breast cancer with oestrogen alone, in women aged under 60 ³.

If a clinician quotes a percentage increase without a baseline, ask for the baseline.

The WHI, in context

The Women's Health Initiative is why a generation of women stopped HRT, and it deserves accurate treatment rather than either dismissal or reverence.

It is the largest randomised trial of hormone therapy in women aged 50 to 79, and the 2022 position statement gave it prominent consideration while being explicit about its limits: the WHI used one route, oral; one estrogen, conjugated equine estrogens at 0.625 mg; and one progestogen, medroxyprogesterone acetate at 2.5 mg. Enrolment of younger symptomatic women was limited, and the trials did not include women with early or premature menopause ³.

The long-term mortality results are the part least often reported. Across 27,347 randomised women followed for a cumulative 18 years, with 7,489 deaths, all-cause mortality was 27.1% in the hormone therapy group and 27.6% in the placebo group, a hazard ratio of 0.99. Cardiovascular mortality, total cancer mortality and mortality from other causes were all statistically indistinguishable between groups ⁴.

A signal of reduced mortality in women aged 50 to 59 was seen during the intervention phase, but the age trend was attenuated and no longer statistically significant over the full 18 years ⁴. The authors were explicit that their findings did not support using hormone therapy to prevent chronic disease or reduce mortality ⁴.

NICE arrives at the same headline for patients: taking either combined or oestrogen-only HRT is unlikely to affect life expectancy ¹.

That is the sentence to take away. Hormone therapy is a treatment for symptoms, with a real and quantified risk profile, that does not appear to shorten or lengthen life at the population level.

What else the tables say

Some outcomes get less attention than they deserve.

Fragility fracture risk is decreased while taking HRT, with the benefit maintained during treatment, decreasing once treatment stops, and potentially continuing longer in people who take it for longer ¹.

Coronary heart disease risk does not increase with either combined or oestrogen-only HRT, and cardiovascular mortality does not increase ¹. This is a meaningful correction to the 2002 headlines.

The risk of developing type 2 diabetes does not increase with HRT, whether oral or transdermal, and generally no adverse effect on blood glucose control is reported ¹.

Ovarian cancer risk increases very slightly, with the baseline population risk in women under 60 described as very low ¹.

Dose, review, and stopping

If a person chooses to take HRT, NICE says use the lowest effective dosage ¹.

Review each treatment at three months to assess effectiveness and tolerability, then annually, unless there is a clinical reason for an earlier review such as ineffectiveness, side effects or adverse events ¹. Refer to a clinician with expertise in menopause if treatments do not improve symptoms or side effects continue ¹.

On stopping, NICE offers a choice of gradually reducing or stopping immediately, and says gradual reduction may limit recurrence of symptoms in the short term but makes no difference in the longer term ¹.

Systemic HRT should be stopped in people diagnosed with breast cancer ¹.

Early and premature menopause is a different calculation

If you reached menopause young, do not read the above as applying to you unchanged.

For premature ovarian insufficiency under 40, NICE recommends offering sex steroid replacement with a choice of HRT or a combined hormonal contraceptive unless contraindicated, and emphasises the importance of continuing until at least the age of natural menopause. It notes that baseline population risk of conditions such as breast cancer and cardiovascular disease increases with age and is very low under 40, and that both HRT and combined oral contraceptives offer bone protection ¹.

For early menopause between 40 and 44, NICE says the benefits and risks of taking or not taking HRT are likely to lie between those for premature ovarian insufficiency and those for people aged 45 or over ¹.

Guidance in the US and Europe

In the US, the FDA announced in November 2025 that it would remove cardiovascular disease, breast cancer and probable dementia from the boxed warning on hormone therapy products, while keeping the boxed warning about endometrial cancer for systemic estrogen-alone products ⁹. The first label changes were approved in February 2026 ¹⁰. The FDA advises starting systemic hormone therapy within 10 years of menopause onset or before 60 ⁹.

That is about treating symptoms. On prevention, the US Preventive Services Task Force still recommends against using hormone therapy to prevent chronic conditions in postmenopausal people without symptoms, finding no net benefit ¹¹.

In Europe, the European Menopause and Andropause Society (EMAS) welcomed the FDA decision. It says that for healthy women under 60 or within 10 years of menopause, the benefits of appropriately chosen hormone therapy generally outweigh its risks, and that treatment must remain individualised rather than one-size-fits-all ¹².

What is still uncertain

The route question, as above: randomised comparisons of oral versus transdermal for hard cardiovascular and cancer outcomes have not been done ³.

Whether progestogen type changes breast cancer risk: NICE says the evidence is insufficient to establish this ¹.

Long-term safety of low-dose vaginal oestrogen beyond one year: the Menopause Society notes that progestogen therapy is not required with low-dose vaginal estrogen but that randomised trial data are lacking beyond a year ³.

And the practical question of how long to hold a dose before judging it, which NICE answers structurally with the three-month review ¹ rather than physiologically.

When to see a clinician

The NHS suggests talking to a GP about HRT if you have menopausal symptoms and want to know your options ⁷. NICE requires clinicians to discuss the benefits and risks, including transdermal versus oral, and to tailor treatment to your age, personal circumstances and risk factors ¹. It also directs them to seek specialist advice on the choice of HRT if you have a condition that may be affected by it ¹.

NICE allows for a review before the annual one if treatment is not working or causes side effects, and recommends referral to a clinician with expertise in menopause if symptoms do not improve ¹.

NICE advises seeking medical help promptly for unscheduled vaginal bleeding outside the expected windows described in the bleeding article on this site ¹.

A note on the record

The guidelines are built around a three-month review ¹. That review works only if someone can say what the past three months were like.

Whatever you use, keep the dates: when you started, when each dose or preparation changed, what improved, what did not, and any side effect with the week it appeared. Halcyon exists to make that record easy to keep, but the guideline does not care what you write it in.