HRT Side Effects And Risks

Most of what the public believes about hormone therapy risk comes from one trial reported in 2002, summarised in relative percentages, and read as though those percentages were absolute. The gap between "a 24% increase in breast cancer" and "8 additional cases per 10,000 women per year" is the entire story of why this treatment became controversial.

This page gives the absolute figures, separates the risks that belong to the hormone from the risks that belong to the delivery route, and covers who should not take it. It is informational; the decision belongs with a prescriber who knows your history.

What the Women's Health Initiative actually found

The Women's Health Initiative remains the largest randomised trial of hormone therapy, and its combined estrogen-plus-progestogen arm was halted early in 2002. The table below gives its principal findings from that 2002 report as absolute event rates, which is how they should have been reported at the time.

Two features of this trial matter as much as the numbers. First, participants had a mean age of 63, and a large proportion were more than a decade past menopause — not the population that typically starts hormone therapy for symptoms. Second, the trial used one specific combination: oral conjugated equine estrogen with medroxyprogesterone acetate. Neither the route nor the progestogen is what most women are prescribed today, and both variables independently affect the risk profile.

Why the route matters more than the dose

The clot and stroke risk associated with hormone therapy is substantially a property of oral administration, not of estrogen itself. An oral tablet is absorbed through the gut and passes through the liver at high concentration before reaching general circulation. That first pass increases hepatic production of clotting factors.

Transdermal delivery — a patch, gel, or spray — enters the bloodstream through the skin and skips that step entirely. Available data has not shown a meaningful increase in venous thromboembolism at standard transdermal doses, which makes this the single most consequential prescribing decision in the whole category.

The progestogen is the second variable. A progestogen is required for any woman with a uterus, because unopposed estrogen thickens the endometrium and raises endometrial cancer risk — that is not a marginal effect and it is why estrogen-alone regimens are reserved for women who have had a hysterectomy. Observational data suggests micronized progesterone may carry a lower breast cancer signal than older synthetic progestins, though this comes from cohort studies rather than randomised trials and should be weighted accordingly. HRT vs BHRT covers the formulation question in detail.

The timing hypothesis

When therapy starts appears to change its cardiovascular effect. The timing hypothesis holds that estrogen started close to menopause acts on relatively healthy arteries, while estrogen started a decade or more later acts on arteries that already carry established plaque — where the effect may be neutral or harmful rather than protective.

This is the most plausible explanation for why the WHI's older cohort showed a coronary signal that trials in newly menopausal women have not reproduced. It is also why the Menopause Society's current hormone therapy position statement distinguishes sharply between initiating therapy before 60 or within ten years of the final period, and initiating it later. It does not mean therapy must stop at 60. It means the decision to start and the decision to continue are different decisions with different evidence behind them.

Common side effects, and which ones need investigating

Most early effects are dose-related and settle within three months. Breast tenderness, bloating, nausea, headaches, and mood changes are the usual list, and persistent breast tenderness more often responds to a dose reduction than to stopping altogether.

Irregular bleeding is the most common reason women discontinue, and it is also the effect where the timeline matters most. Unscheduled bleeding in the first six months of a continuous combined regimen is expected as the endometrium adjusts. Bleeding that begins after a period of stability, or that persists beyond six months from initiation, falls into a different category and warrants investigation rather than reassurance. So does any bleeding that starts before therapy begins — that needs working up first, not after.

Who should not take hormone therapy

Absolute contraindications are current or past breast cancer, other estrogen-dependent cancers, unexplained vaginal bleeding not yet investigated, active or recent venous thromboembolism, active or recent arterial events including heart attack and stroke, active liver disease, and pregnancy.

Relative contraindications need individual assessment rather than an automatic no: a strong family history of breast cancer, a known inherited thrombophilia, migraine with aura, uncontrolled hypertension, and gallbladder disease. Several of these push the decision toward transdermal rather than oral delivery rather than away from therapy entirely — which is precisely the kind of nuance a blanket risk headline erases.

Monitoring is the other half of the picture. Blood pressure, a symptom review, and a discussion of whether the current dose and route are still right belong in an annual review, alongside age-appropriate breast screening. A perimenopause biomarker panel covers what is worth measuring before and during treatment, and doctor-led options lists the physician-supervised routes for getting that review.

If systemic therapy is off the table, what is left

A contraindication to systemic hormone therapy is not a contraindication to treatment. Three routes remain, and the first is the one most often missed because it sits in the table above without ever being connected to the contraindication list.

Vaginal estrogen is a separate decision. At standard low doses it acts locally with minimal systemic absorption, so the systemic risk profile that rules out a patch or tablet does not transfer to it. For vaginal dryness, painful sex, urinary urgency, and recurrent UTIs it is often available to women who cannot take systemic therapy. Breast cancer survivors — particularly anyone on an aromatase inhibitor — are the exception that needs an oncologist in the conversation rather than a general prescriber.

Non-hormonal prescriptions treat hot flushes directly. Low-dose paroxetine is FDA-approved for moderate-to-severe vasomotor symptoms; venlafaxine, escitalopram, and gabapentin are used off-label with randomised support behind them. Fezolinetant, a neurokinin-3 receptor antagonist approved in 2023, works through a different mechanism entirely and requires liver-function monitoring, which the prescriber should set up rather than leave to you. None of these matches hormone therapy for symptom control, but each meaningfully reduces frequency and severity.

Cognitive behavioural therapy has trial evidence for menopausal symptoms, and it works on the distress and sleep disruption rather than the flush itself — which for many women is the part that actually degrades daily life. It is also the only option on this page with no contraindications at all. Raise all three with a prescriber together; the common failure is being told "you can't take HRT" and hearing it as the end of the conversation.

Weighing it honestly

Hormone therapy has documented benefits alongside its risks, and a risk page that omits them misinforms as badly as a marketing page that omits the risks. It is the most effective available treatment for vasomotor symptoms, it treats genitourinary symptoms that do not improve on their own, and it reduces fracture risk consistently across trials — that hip fracture line in the WHI table is a real benefit measured in the same trial as the harms.

The reasonable framing is not "is HRT safe" but "for me, at my age, with my history, does this specific formulation and route do more good than harm — and is that still true in a year?" That question has an answer. It just is not the same answer for everyone, and it is not one a website can give you.

Hormone Therapy After A Breast Cancer Diagnosis

Systemic hormone therapy is generally avoided after a breast cancer diagnosis, and the reasoning is strongest for oestrogen-receptor-positive disease. That leaves a group with severe menopausal symptoms, often made worse by the treatment itself, and a treatment they cannot use. Aromatase inhibitors and ovarian suppression both produce vasomotor symptoms more intense than natural menopause does.

Four non-hormonal options have evidence behind them, and one of them is newer than most patient material acknowledges.

• SSRIs and SNRIs. Venlafaxine and escitalopram both reduce hot flush frequency in trials. Paroxetine and fluoxetine inhibit the CYP2D6 enzyme that converts tamoxifen to its active form, so they are avoided in anyone taking tamoxifen.
• Gabapentin. Useful where night sweats dominate, since sedation works in your favour at bedtime.
• Fezolinetant. A neurokinin 3 receptor antagonist approved by the US Food and Drug Administration in 2023 for moderate to severe vasomotor symptoms. It acts on the brain pathway that generates hot flushes rather than on oestrogen receptors, which is why it is available to women who cannot take hormones. Liver monitoring is required.
• Cognitive behavioural therapy for menopausal symptoms. It reduces how much the symptoms interfere with daily life rather than the flush count, and it has randomised evidence in breast cancer survivors specifically.

Vaginal symptoms are handled separately. Low-dose vaginal oestrogen has minimal systemic absorption, and its use in breast cancer survivors with severe genitourinary symptoms is a decision some oncology teams will support after weighing it, particularly where non-hormonal moisturisers have failed. That conversation belongs with the oncologist holding your pathology rather than with a menopause platform.

Related

• Women's health guides — menopause, hormones, and what to measure
• HRT vs BHRT — what the formulation difference actually is
• The three stages of menopause
• Perimenopause biomarker panel
• Doctor-led options — physician-supervised routes, compared