Hormone therapy: what actually changed after the WHI

In July 2002, the Women's Health Initiative announced it was halting one arm of a large randomised trial of hormone therapy early. The headlines said HRT caused breast cancer. Prescribing fell by more than half within a year, in some countries by considerably more, and an entire cohort of women went untreated.

What the trial actually found is more nuanced, and the subsequent two decades of reanalysis have changed the picture substantially. This is worth understanding, because it is still the reason a lot of doctors are hesitant.

What the WHI actually studied

The trial randomised postmenopausal women to hormone therapy or placebo. Two arms: conjugated equine oestrogens plus a progestin (medroxyprogesterone acetate) for women with a uterus, and oestrogen alone for women who had had a hysterectomy.

Three design features matter enormously for how the results should be read:

The average age at enrolment was 63. Over two-thirds of participants were more than ten years past menopause. This was not a trial of women with menopausal symptoms starting treatment at 51 — it was largely a trial of primary prevention in older women.

The formulations were of their era. Oral conjugated equine oestrogens and medroxyprogesterone acetate. Not transdermal oestradiol, not micronised progesterone, both of which are now widely used and have different risk profiles.

The primary outcome was cardiovascular prevention, not symptom relief.

What it found

In the combined oestrogen-plus-progestin arm: a small increase in breast cancer diagnoses, an increase in stroke and venous thromboembolism, and no cardiovascular benefit. The absolute increase in breast cancer was on the order of a few extra cases per thousand women per year — a real signal, and one that was widely reported as a relative risk without the absolute numbers, which made it sound far larger.

In the oestrogen-only arm — the part that received almost no coverage — there was no increase in breast cancer, and in longer follow-up a reduction in breast cancer incidence and mortality. That finding alone should have complicated the headline considerably.

What changed since

The timing hypothesis. Subsequent analyses, including of the WHI's own data by age subgroup, found that women who started hormone therapy within about ten years of menopause or before age 60 had a different risk-benefit profile — including a possible reduction in all-cause mortality — compared with those starting much later. Starting close to menopause appears to behave differently from starting in your late sixties.

Route matters. Transdermal oestrogen — patches, gels, sprays — bypasses first-pass liver metabolism and is not associated with the increased venous thromboembolism risk seen with oral preparations. This is now a standard consideration in prescribing.

Progestogen type matters. Micronised progesterone appears to carry a lower breast cancer signal than older synthetic progestins, though the comparative data is observational rather than from head-to-head trials.

The absolute risks are small and comparable to familiar things. Major guidelines now note that the breast cancer risk associated with combined HRT is broadly in the range of that associated with drinking a couple of units of alcohol a day, or with being physically inactive — comparisons that are rarely offered.

Where guidance sits now

Major bodies — including the North American Menopause Society and NICE — currently hold that for symptomatic women under 60 or within ten years of menopause, the benefits of hormone therapy generally outweigh the risks. It is the most effective treatment for vasomotor symptoms, and it protects bone.

There is no arbitrary time limit on duration; the recommendation is periodic review rather than a fixed stopping point.

For primary ovarian insufficiency or early menopause, hormone therapy is recommended until the average age of natural menopause, and the framing is replacement of what should be there rather than treatment of symptoms. That is a different risk calculation entirely.

Vaginal oestrogen is a separate category. Systemic absorption is minimal, it is not associated with the risks above, and it is appropriate for almost everyone with genitourinary symptoms — including many women who cannot take systemic HRT. It is drastically under-prescribed. Vaginal and urinary changes after menopause.

Who should not, or should be careful

Current or recent breast cancer, oestrogen-dependent cancers, unexplained vaginal bleeding, active liver disease, current venous thromboembolism or a history of it, and previous stroke or heart attack are contraindications or require specialist input. Migraine with aura affects the choice of route rather than ruling treatment out.

None of this is a conversation to have from an article. It is a conversation to have with a clinician who knows the current guidance.

Non-hormonal options exist and are legitimate

CBT has real evidence for both vasomotor symptoms and the mood side. Certain SSRIs and SNRIs, gabapentin, and newer neurokinin-3 receptor antagonists developed specifically for hot flushes are all options. Hot flushes: mechanism and what helps covers them.

What to bring to the appointment

Hormone therapy decisions are individual, and the two things that make the conversation productive are a clear symptom picture and a clear sense of what you want to change.

That means: which symptoms, since when, how severe, what they are stopping you doing, and what you have already tried. Dated, not remembered. It also means being able to tell afterwards whether it worked — which sounds obvious and is surprisingly hard without a baseline, because symptoms fluctuate and memory adjusts.

Naked is built to hold both ends of that: the months before, so the case is clear, and the months after, so you can see what actually changed. Starting or adjusting HRT is an experiment, and experiments need a baseline.