Menopausal Hormone Therapy (MHT)

Healthcare · §635

Around age fifty-one the ovaries stop, and the estrogen that was running sleep, mood, bone, blood vessels, vaginal tissue, and parts of the brain drops to near zero. Menopausal hormone therapy puts most of it back — and for symptomatic women under sixty, it is one of the highest-value prescriptions in medicine. The 2002 trial that drove two decades of doctors off it enrolled women a median twelve years past menopause; that mismatch between the trial cohort and the symptomatic woman in front of the GP is the whole story behind the 2025 FDA reconsideration of the boxed warning. What follows is when it works, when timing has run out, and which boxes on the warning label were never earned.

Decide · Daily Evidence Moderate Chapter Healthcare

The dominant menopausal symptom — flashes by day, night sweats that fracture sleep for a median seven years — eases by about three-quarters within weeks of an adequate dose. Bone fractures drop while you take it. A separate, low-dose vaginal form reverses the dryness, painful sex, and recurrent infections that follow menopause without putting the hormone into the rest of the body. The real decision is whether the regimen matches your timing, your uterus status, and your cardiovascular history — and finding a doctor who works through that with you instead of handing back the 2002 warning label.

Estrogen is not a sex hormone in the narrow sense — it is one of the body's main signalling molecules, with receptors on bone, blood-vessel walls, the temperature-control circuitry in the hypothalamus, the lining of the vagina and bladder, the parts of the hippocampus that hold short-term memory, and the skin's collagen-producing cells Mehta 2021. When the ovaries wind down, the level in the blood falls by roughly ninety-five percent. The body does not adapt — it runs without the signal, and most of the menopausal experience is what happens to those tissues when the signal goes missing.

The hot-flash mechanism is the cleanest example. A small group of neurons in the hypothalamus controls body temperature; estrogen normally keeps them in check, and when it disappears they fire in unstable bursts that the body reads as too hot and corrects with a sudden flush, sweat, and skin flush. Put the estrogen back and the firing settles within weeks NAMS 2022. The same pattern plays out for vaginal tissue (thins and dries without the signal; rehydrates and thickens with replacement), for bone (the cells that break it down lose their brake, and a year of menopause costs roughly one to three percent of bone mass), and for the blood-vessel lining (loses some of the nitric-oxide signalling that keeps it relaxed).

Replacement comes in two regimens. Women with a uterus need an estrogen plus a progestogen — the second hormone keeps the uterine lining from over-growing, which is the same protection a normal cycle was doing. Women who have had a hysterectomy take estrogen alone, and the absence of the progestogen turns out to matter for the safety profile (more on that below). Both come as oral pills, skin patches, gels, or sprays. A separate, lower-dose version delivered locally — vaginal cream, tablet, or ring — treats the dryness and bladder symptoms without raising blood levels meaningfully above the postmenopausal baseline NAMS GSM 2020.

What the trials actually showed, and what the warning got wrong

Hormone replacement was the most-prescribed drug class in postmenopausal American women through the 1990s, on the back of observational data suggesting it cut heart attacks. Then a single trial reset the field. The Women's Health Initiative randomised over sixteen thousand postmenopausal women to a combined estrogen-and-progestin pill or placebo. In 2002 the trial's safety board halted the combined arm early, citing a small excess of breast cancer, heart attacks, strokes, and blood clots Rossouw 2002. American prescribing fell by eighty percent inside two years. The FDA put a boxed warning on every systemic hormone product, and a generation of primary-care doctors learned to say no.

The trial's authors started saying something else within five years. The women they had enrolled were a median twelve years past menopause and an average age of sixty-three — already inside the window where arteries have laid down plaque. When they re-analysed by age band, the heart-attack signal lived almost entirely in the seventy-plus stratum; women within ten years of menopause showed the opposite trend Rossouw 2007. This is the timing hypothesis: estrogen restarted into a still-healthy artery protects it; estrogen restarted into an artery already carrying unstable plaque can destabilise it, especially in the first year or two.

Three RCTs designed to test the timing hypothesis directly. The Danish DOPS trial randomised about a thousand recently menopausal women to hormone therapy or no treatment and followed them ten years. The treated arm had a roughly fifty-percent reduction in the combined risk of death, heart attack, and heart failure — with no excess breast cancer or stroke Schierbeck 2012. ELITE randomised by time-since-menopause: women within six years showed slowed thickening of the carotid artery wall on oral estradiol; women ten-plus years out showed no benefit, confirming the time-window mechanism Hodis 2016. KEEPS, a four-year trial, showed no harm to artery imaging and improved quality of life in either oral or patch form Harman 2014. A 2015 stratified meta-analysis pooling the early-start trials found a roughly fifty-percent reduction in coronary deaths and heart attacks and a thirty-percent reduction in all-cause mortality Boardman 2015.

The breast-cancer story turned out to depend on which hormones. In the WHI's combined arm (estrogen plus a synthetic progestin called medroxyprogesterone), breast cancer incidence rose modestly and persistently — roughly one extra case per thousand women per year above baseline Chlebowski 2020. In the WHI's estrogen-only arm (women who had had a hysterectomy and didn't need the second hormone), breast cancer incidence fell, and so did breast cancer mortality Anderson 2004 Chlebowski 2020. A worldwide individual-participant meta-analysis of half a million women confirmed both signals and showed the excess scales with how long you take it; the absolute numbers are small for any single woman but real CGHFBC 2019. The clean takeaway most lay coverage misses: which hormones, not hormones.

On bones, the WHI is one of the only trials in medicine that has ever cut hip fractures in a low-risk population. Combined therapy reduced clinical fractures by a quarter and hip fractures by a third; estrogen alone did the same Cauley 2003 Anderson 2004. The protection persists while you take it and fades within a few years of stopping — replacement is treatment, not a one-time fix.

On the brain, the picture splits the same way the heart did. The Women's Health Initiative Memory Study followed women aged sixty-five and older starting combined therapy and found a doubling of dementia risk — the strongest single argument against late initiation Shumaker 2003. The same study's younger cohort, followed seven years on, showed no harm and no benefit to cognition Espeland 2013; the ELITE cognitive sub-study showed the same neutrality in early starters Henderson 2016. The strongest brain signal comes from women whose ovaries were surgically removed before natural menopause: without estrogen replacement through to roughly age fifty-one, they show a measurably higher rate of cognitive impairment decades on Rocca 2007. The brain, like the artery, has a window.

The 2025 FDA panel was the regulatory acknowledgement of all of this. The agency announced reconsideration of the boxed warning that has been class-applied to every systemic hormone product, including the low-dose vaginal preparations that never carried any of the trial risk signals to begin with FDA 2025 Manson 2024. The boxed warning has not been removed at the time of writing; the institutional movement is the news.

What continues if you don't treat

The median woman has hot flashes for seven and a half years. The hardest-hit groups average over ten. That is most of the runway between fifty and sixty: a decade of waking three or four times a night drenched, the meeting interrupted by a flush you can feel climbing your neck, the husband sleeping in the spare room because you've kicked off the duvet again. People who haven't been through it tend to think of it as inconvenience. Women a year in describe it as the floor of their daily life coming out.

The genitourinary changes don't pass with time — they progress. Vaginal dryness becomes painful sex becomes no sex. Urinary urgency becomes recurrent infections becomes the antibiotic-of-the-month problem. By your late sixties, the connective tissue of the pelvic floor has thinned to the point that a third of women have some form of incontinence and that fraction grows from there NAMS GSM 2020. None of this is inevitable; all of it reverses with local estrogen, which the literature does not show any meaningful systemic risk profile for.

Bone is the silent one. You don't feel the one-to-three percent of bone mass leaving your spine and hips per year through the menopausal transition. You feel it for the first time when you trip on a curb at sixty-eight and shatter a wrist, or step off a kerb wrong at seventy-two and fracture a hip — and the one-year mortality after a hip fracture in a woman past seventy-five is roughly twenty percent. Hormone replacement, taken through the transition window, is one of the few interventions that has reduced hip fractures in a primary-prevention trial Cauley 2003.

And in the corner of the population where menopause came early or surgically — ovaries removed before forty-five for a cyst or a cancer scare, an early natural menopause at forty — the long-run picture without replacement is a measurably higher rate of cardiovascular disease and a measurably higher rate of cognitive impairment in your seventies Rocca 2007. Not treating premature menopause is its own choice with its own data.

How it's actually taken

The 2022 North American Menopause Society position is the field's reference document; the Endocrine Society guideline reads nearly identically. Both frame hormone therapy as first-line for symptomatic women under sixty or within ten years of menopause, with no contraindications, in a shared decision-making conversation with the prescriber NAMS 2022 Stuenkel 2015. The shape of the prescription itself is well-mapped.

The estrogen. Oral estradiol 0.5 to 1 milligram a day, or a transdermal patch at 25 to 100 micrograms a day changed twice a week, or the gel/spray equivalents. Patch is preferred if you have a higher clot risk — heavier body, family history of clots, migraine with aura, gallbladder trouble — because it skips the liver pass that raises clotting factors Canonico 2007.
The progestogen, if you have a uterus. Micronized progesterone 100 milligrams nightly with continuous estrogen, or 200 milligrams for twelve nights a month if you're still cycling. The hormonal IUD is an off-label but increasingly common alternative.
The vaginal estrogen, if that's your problem. Cream, tablet, or ring at one-tenth or less the systemic dose. Use it alone or alongside systemic therapy. The blood levels don't rise above the postmenopausal baseline NAMS GSM 2020.
The first three months. Adjust dose to symptom relief. If hot flashes are still breaking through at twelve weeks, go up — under-dosing is a more common error than over-dosing.
The annual. Mammogram and breast exam, blood pressure, the usual cardiovascular labs, a check of how things are going. No need to recheck blood levels routinely.
Duration. NAMS 2022 explicitly threw out the five-year-and-stop rule. Continue as long as the symptoms warrant and the risk-benefit still favours continuing NAMS 2022.

What it costs and how to find a prescriber

Generic oral estradiol and the transdermal patch run roughly two to six hundred dollars a year in the US with insurance; branded combinations and vaginal rings climb from there. Without insurance the same generics are still affordable from major pharmacies. The progesterone component is similarly cheap. A daily pill or a twice-weekly patch — the dosing itself is the easiest part of any prescription you will ever take.

The friction sits in the prescriber. Many primary-care doctors trained during the post-2002 fear era still default to declining or hand back the boxed warning. A NAMS-certified menopause practitioner — the certification list is public on the Menopause Society's site — is the more reliable route, sometimes via self-referral and sometimes self-pay. Telehealth menopause clinics have emerged to fill the gap and can be a workable entry point if local options are thin Manson 2024.

When not to take it

Hormone therapy is prescription-only because the contraindication list is real and the routing decisions (oral vs patch, which progestogen, vaginal vs systemic) depend on personal history a friend on the internet cannot see.

Personal history of breast cancer or other estrogen-sensitive cancer. Systemic estrogen is generally avoided; some breast cancer survivors do use vaginal estrogen for severe genitourinary symptoms after a conversation with their oncologist NAMS GSM 2020.
History of blood clots, stroke, or heart attack. Systemic therapy is contraindicated; vaginal estrogen is a separate question.
Active liver disease, unexplained vaginal bleeding, or known pregnancy. Don't start.
You're more than ten years past menopause or over sixty, with no symptoms. The timing window for starting fresh has likely closed; the cardiovascular and dementia risk-benefit shifts away from initiation Rossouw 2007 Shumaker 2003.
Relative cautions. Heavier body, untreated high blood pressure, migraine with aura, high triglycerides, gallbladder disease, family history of clots — all point toward the patch rather than the pill, not toward avoiding hormones Canonico 2007.

Vaginal estrogen has a much narrower exclusion list — the boxed warning that has been class-extended to local preparations was never derived from trial data on those products and is part of what the 2025 FDA reconsideration is about FDA 2025.

The myths that kept a generation of women off it

"It causes breast cancer." Estrogen plus a synthetic progestin slightly raises breast cancer risk — about one extra case per thousand women per year on top of baseline, accumulating with how many years you take it Chlebowski 2020. Estrogen alone, in women who have had a hysterectomy, reduces breast cancer incidence and breast cancer mortality Anderson 2004. Lay coverage almost never makes the distinction.

"It causes heart attacks." Started within ten years of menopause, hormone therapy reduces coronary events and overall mortality. Started twenty years out, in arteries that already have established plaque, it can trigger events in the first year or two Rossouw 2007 Hodis 2016 Boardman 2015. The same drug, the same dose; the difference is the artery.

"You have to stop at five years." NAMS 2022 explicitly rejects this rule. Duration is whatever the ongoing risk-benefit supports — there is no biological clock that triggers harm at year six NAMS 2022.

"Bioidentical compounded hormones are safer than the FDA-approved kind." Compounded preparations don't have dose verification, are not held to the same purity standards, and have been associated with cases of uterine over-growth in surveillance reports because no one was sure how much progesterone the cream was actually delivering NAMS 2022. The FDA-approved estradiol and patches are bioidentical to what the ovary made; the compounded marketing relies on the word, not the molecule.

"Vaginal estrogen carries the same risks as the pill." Low-dose vaginal estrogen produces blood levels at or below the postmenopausal baseline NAMS GSM 2020. The warning label on the box was extended from the systemic trials to the vaginal products with no supporting trial evidence — which is one of the items the 2025 FDA panel is reconsidering FDA 2025.

"Hot flashes pass; just ride it out." Median duration is over seven years, longer for women whose symptoms started early or who are Black NAMS 2022. Riding it out is a real choice; framing it as the easy default is misleading.

Who specifically

This entry is written for women in or past the menopausal transition. Within that, four subgroups deserve being named.

Symptomatic women in their fifties, within ten years of menopause. The strongest case. Hot flashes, night sweats, sleep fragmentation, mood lability, brain fog — the cluster of symptoms that defines perimenopause for most. The risk-benefit balance for systemic hormone therapy is favourable in this window; the heart and brain trial data are reassuring; the symptom relief is roughly three-quarters of flashes gone within weeks MacLennan 2004 Schierbeck 2012. If you are this reader, the question to your prescriber is not whether but which route and what dose.

Women whose ovaries stopped early — surgically or naturally before forty-five. All major guidelines recommend hormone replacement at least through age fifty-one in this group. Withholding is associated with measurably higher rates of cardiovascular disease and cognitive impairment decades on Rocca 2007 NAMS 2022. This is the population where not treating is the active choice with its own risk profile.

Women past sixty with genitourinary syndrome only. Vaginal dryness, painful sex, recurrent urinary infections, urgency. Low-dose vaginal estrogen — cream, tablet, or ring — reverses these without raising blood hormone levels meaningfully. The systemic risk-benefit conversation does not apply, the contraindication list is much narrower, and the relief is reliable within weeks NAMS GSM 2020. The class warning on the package was never derived from these products.

Women past sixty who are already on systemic hormone therapy and well. NAMS 2022 explicitly supports continuation when the symptom benefit is real and the risk profile has stayed favourable. There is no biological reason to stop just because of the calendar NAMS 2022.

Two groups for whom the answer is generally different: women more than ten years past menopause with no current symptoms (the timing window has likely closed for initiating systemic therapy), and women with a personal history of breast or other estrogen-sensitive cancer (specialist decision territory, often with vaginal estrogen as the conditional yes after a conversation with the oncologist).

Why "I tried it and it didn't work" usually has a specific cause

Dose too low. The most common error. Standard starting estradiol of half a milligram clears flashes for some women and leaves others still drenched at week eight. Titrate up to a milligram and beyond if symptoms persist — under-dosing for fear of side effects is the modal failure.
Wrong route for your risk profile. A heavy woman with a family history of clots put on oral estrogen will accumulate clotting risk that a transdermal patch would have skipped Canonico 2007. The patch is a routing decision your prescriber should make explicitly.
Wrong progestogen. Medroxyprogesterone (the synthetic in the original 2002 trial) is the one with the breast-cancer signal and the mood side effects. Micronized progesterone is the bioidentical option most modern menopause prescribers reach for; the hormonal IUD is the third NAMS 2022.
Initiated too late. Starting hormone therapy fresh at age sixty-eight in a woman whose arteries are already lined with plaque is the scenario the original heart-attack signal lives in Rossouw 2007 Hodis 2016. The timing matters.
Compounded preparations. Marketed as "bioidentical" but without dose verification or purity testing. The progesterone component is often under-delivered, which is the surveillance signal for uterine over-growth NAMS 2022.
Treating only the systemic symptoms when the vaginal ones are the real problem. Systemic estrogen helps the vaginal tissue, but local low-dose vaginal therapy gives the biggest, fastest relief for dryness, painful sex, and recurrent infections — and a lot of women never get offered it because the prescriber went straight to the pill NAMS GSM 2020.
Stopping cold turkey. A sudden withdrawal often produces a hot-flash rebound worse than what you started with, and Finnish national-registry data flags an excess of cardiac deaths in the first year after abrupt withdrawal in women under sixty Mikkola 2015. Taper.

What changes, and when

Honest about timing — different effects show up on different clocks.

Two to four weeks. Hot flashes start dropping. By week six on an adequate dose, the version of you that was waking three times a night drenched is the version that sleeps through and wakes once, to a calm room. Your partner notices first; you notice the morning after.

Four to twelve weeks. Vaginal tissue rehydrates and thickens on local or systemic estrogen. Sex stops hurting. The recurrent urinary infections that started showing up in your fifties stop showing up. Skin holds moisture longer; the chronic dryness behind your knees and on your shins fades NAMS GSM 2020.

Three to six months. Mood steadies. The perimenopausal irritability — the snapping at the kids you weren't snapping at last year, the day-shaped-by-a-bad-Monday — comes back under your control. The fog clears, mostly because you are sleeping again. The conversations you were postponing because you couldn't quite hold the thread are conversations you finish.

One to two years. Bone density measurements stabilise or begin to recover. The trajectory you were on — losing one to three percent of bone a year through the menopausal transition — has flattened Cauley 2003. You will not feel this; the X-ray report will.

Decade-scale. If you started early, this is where the cardiovascular and mortality numbers sit. The Danish trial that followed recently-menopausal women on hormones for ten years measured a roughly fifty-percent reduction in the combined risk of death, heart attack, and heart failure compared to controls Schierbeck 2012. The Bayesian meta-analysis of younger-starter trials lands around a forty-percent all-cause mortality reduction Salpeter 2009. The eighteen-year WHI follow-up sits more neutrally — no benefit and no harm in the long-run mortality average Manson 2017. The honest summary: the signal is real and probably positive in early starters, but a hard-endpoint trial specifically powered for cardiovascular death in the early-postmenopausal woman has not been run.

The premature-menopause case. For a woman whose ovaries stopped in her thirties or early forties, replacement through to about age fifty-one is what the trajectory was supposed to look like. The mortality and cognitive numbers in this group, without replacement, are measurably worse than in the general population Rocca 2007. With replacement, the difference is largely closed.

Adjacent topics worth knowing about: the non-hormonal hot-flash drugs (low-dose antidepressants like paroxetine or venlafaxine used off-label, gabapentin, and the newer drug fezolinetant that targets the same hot-flash circuit without hormones) for women who can't or won't take estrogen; resistance training and adequate dietary protein, which are the rest of the bone-preservation story alongside or in place of hormone therapy; vitamin D and calcium status, the same; cardiovascular risk testing in your fifties (lipid panel, blood pressure, ApoB) so the decision about which route is informed by data rather than guesswork; and pelvic floor physiotherapy for the urinary symptoms that don't fully resolve on local estrogen alone.

Related in the handbook

Helps

— HRT is aimed squarely at the hot flashes, sleep, and mood swings this transition map lays out.
— Beyond hot flashes, estrogen therapy slows the bone loss that leads to the fractures osteoporosis care is racing.

— Vaginal dryness and recurrent UTIs often need local estrogen even if you skip systemic HRT — a separate, lower-dose tool.
— Menopause and its treatment sit inside the broader picture of women's heart risk.
— Strength training is the behavioural partner to HRT for holding bone and muscle through the transition.
— Risk-reducing surgery for BRCA carriers triggers early menopause, making the HRT decision part of the plan rather than an afterthought.
— The 2002 scare that drove women off HRT was about breast cancer; keeping up with screening makes the decision clearer, not scarier.
— In perimenopause, contraception and symptom relief overlap — a hormonal IUD often bridges both before full menopause therapy.
— Migraine with aura changes the hormone-safety math — it's a key question before starting any estrogen.

Substance and claimed effects

Menopausal hormone therapy (MHT, historically called HRT) replaces the estrogen — and, when the uterus is intact, the progestogen — that ovarian failure withdraws around age 51. The pharmacological backbone is estradiol or conjugated equine estrogen, delivered orally, transdermally (patch / gel / spray), or vaginally (cream / tablet / ring), paired with a progestogen (oral micronized progesterone, levonorgestrel IUD, or a synthetic progestin) in women with a uterus to oppose endometrial proliferation. Claimed and evidence-supported effects, all in scope for this entry: substantial relief of vasomotor symptoms (hot flashes, night sweats) — the dominant indication MacLennan 2004; reversal of the genitourinary syndrome of menopause (vaginal dryness, dyspareunia, recurrent UTI, urgency) via systemic or low-dose vaginal estrogen NAMS GSM 2020; prevention of osteoporotic fracture Cauley 2003; modulation of cardiovascular disease risk that depends critically on timing of initiation — the timing hypothesis: protective if started within ~10 years of menopause or before age 60, neutral-to-harmful if started later Rossouw 2007 Hodis 2016 Manson 2017; breast cancer risk modulation that diverges sharply by regimen — estrogen alone (in hysterectomized women) reduces incidence, estrogen plus progestin increases it modestly Chlebowski 2020; sleep, mood, and quality-of-life improvements largely secondary to vasomotor relief; and cognitive outcomes that hinge on age of initiation — neutral or possibly protective near menopause, harmful when started in women aged 65+ Shumaker 2003 Henderson 2016. The FDA in 2025 convened an expert panel and announced reconsideration of the boxed warning that has sat on systemic MHT products since 2003 — the regulatory acknowledgement that the WHI's class-warning approach, derived from women a median 12 years past menopause, has been applied to women for whom it was never appropriate FDA 2025 Manson 2024.

Evidence by addressing question

mechanism

Estradiol is the dominant ovarian estrogen during reproductive life. Ovarian follicular depletion at menopause drops circulating estradiol from a cycling range of ~50–400 pg/mL to ~5–20 pg/mL. Estrogen receptors (ERα, ERβ, and the membrane GPER) sit on virtually every tissue: hypothalamic thermoregulatory neurons in the preoptic area, urogenital epithelium, osteoblasts and osteoclasts, vascular endothelium and smooth muscle, hippocampal and prefrontal neurons, mammary epithelium Mehta 2021. Withdrawal produces the canonical menopausal syndrome via three mechanisms: dysregulation of the hypothalamic temperature set-point (driven by hyperactivity of KNDy — kisspeptin/neurokinin B/dynorphin — neurons unopposed by estrogen), producing vasomotor symptoms; atrophy of estrogen-dependent epithelia (vaginal, urethral, trigonal), producing GSM; and an accelerated phase of bone resorption as estrogen restraint on osteoclast activity is lifted, producing 1–3% annual bone loss for the first 5–10 postmenopausal years NAMS 2022.

Cardiovascular mechanism is regimen- and timing-dependent. In recently menopausal arteries, estradiol promotes endothelial nitric oxide, improves lipid profile (LDL ↓, HDL ↑), and inhibits early atherosclerotic plaque formation — the basis of the "healthy endothelium hypothesis." In arteries already laden with established, possibly unstable plaque (typical of women a decade or more past menopause), estrogen's prothrombotic and pro-inflammatory effects on vulnerable plaque can precipitate events, especially in the first 1–2 years of treatment Hodis 2016 Rossouw 2007. Oral estrogen is metabolized through the liver and raises clotting factors and triglycerides; transdermal estrogen bypasses first-pass hepatic metabolism and confers minimal-to-no excess venous thromboembolism risk in observational data Canonico 2007. The progestogen component matters: medroxyprogesterone acetate (the WHI synthetic, MPA) appears to drive most of the combined-arm's breast cancer signal and may oppose some of estrogen's vascular benefit; micronized progesterone and the levonorgestrel IUD have more favourable profiles in observational data, though no head-to-head mortality RCT exists NAMS 2022 CGHFBC 2019.

evidence

Vasomotor symptoms. The most settled signal in the literature. Cochrane meta-analysis of 24 placebo-controlled RCTs found systemic estrogen reduced hot-flash frequency by ~75% (weighted mean ~18 fewer flushes per week) and severity by similar magnitudes versus placebo MacLennan 2004. Effect sizes are larger than any non-hormonal alternative (SSRIs, gabapentin, oxybutynin, fezolinetant) on a head-to-head basis Stuenkel 2015 NAMS 2022.

Bone. WHI estrogen-plus-progestin reduced clinical fractures by HR 0.76 (95% CI 0.69–0.85) and hip fractures by HR 0.67 — the only therapy that has reduced fractures in a low-risk population in a primary-prevention RCT Cauley 2003. The estrogen-alone arm produced similar reductions Anderson 2004. The effect attenuates within 2–5 years of stopping, so MHT is bone-protective while taken and largely not after.

Genitourinary syndrome of menopause. Low-dose vaginal estrogen (cream, tablet, ring) restores epithelial thickness, vaginal pH, and lubrication; reduces dyspareunia and recurrent UTI; with serum estradiol levels at or below the postmenopausal baseline. NAMS 2020 GSM position statement: vaginal estrogen carries none of the systemic-MHT risk profile and the boxed warning class-extended to vaginal products has never had supporting trial data NAMS GSM 2020.

Cardiovascular — the timing hypothesis. WHI E+P (n=16,608, mean age 63, median 12 years past menopause) reported HR 1.29 for coronary heart disease, prompting the 2002 trial halt and the boxed warning Rossouw 2002. Subsequent age-stratified re-analysis showed the CHD signal was driven entirely by women in the 70–79 age stratum; women within 10 years of menopause had HR 0.76 (non-significant) Rossouw 2007. The Danish DOPS RCT (n=1006, recently menopausal women, 10-year follow-up) reported a 52% reduction in the composite of death, MI, and heart failure (HR 0.48, 95% CI 0.26–0.87) with no excess cancer or stroke Schierbeck 2012. ELITE (n=643) randomized women by stratified time-since-menopause; the early-stratum (<6 years) showed slowed carotid intima-media thickness progression with oral estradiol vs placebo; the late-stratum (≥10 years) showed no benefit — the strongest mechanistic confirmation of timing Hodis 2016. KEEPS (n=727, oral CEE or transdermal estradiol, 4 years) found no atherosclerosis progression difference vs placebo in any arm but improved quality of life and lipid markers Harman 2014. Cochrane 2015 stratified meta-analysis found women starting MHT <10 years from menopause had a 48% reduction in coronary deaths and MI (RR 0.52, 95% CI 0.29–0.96) and 30% all-cause mortality reduction Boardman 2015. WHI 18-year cumulative follow-up: no statistically significant difference in all-cause or cause-specific mortality versus placebo across either arm, but in women initiating at 50–59 the point estimate trended favourable for all-cause mortality Manson 2017.

Breast cancer — regimen-dependent. In WHI's estrogen-plus-progestin arm: HR 1.24 for invasive breast cancer (5.6 years intervention), persistent at long-term follow-up but with HR for breast-cancer mortality 1.45 — modest absolute increase (~1 extra case per 1000 women per year over baseline) Chlebowski 2020. In the estrogen-alone arm (hysterectomized women, conjugated equine estrogen): HR 0.78 for breast cancer incidence and HR 0.60 for breast cancer mortality — a statistically significant reduction Anderson 2004 Chlebowski 2020. The CGHFBC individual-participant meta-analysis (n>500,000) replicated the directional finding: 5 years of E+P from age 50 produces ~8.3% lifetime breast cancer risk vs 6.3% in never-users (an excess of ~1 per 50 over 20 years); estrogen-alone produces a smaller excess (~1 per 200) CGHFBC 2019. The Million Women Study reported a larger E+P risk (RR ~2.0 for current users) but is observational and likely biased by detection and confounding Beral 2003. The author's call below addresses the gap.

Cognition. WHIMS (substudy of WHI, women aged 65+ at initiation): E+P doubled dementia incidence (HR 2.05) and increased mild cognitive impairment — the strongest argument against late initiation Shumaker 2003. WHIMSY (women initiated in WHI at 50–55, cognition assessed at 7-year follow-up): no harm and no benefit on cognition Espeland 2013. ELITE-Cog (early- vs late-onset estradiol, 5-year cognitive battery): neutral in both strata Henderson 2016. Mayo cohort: women who underwent bilateral oophorectomy before age 45 without estrogen replacement showed a hazard ratio of ~1.7 for cognitive impairment and dementia compared with referents — the strongest evidence that estrogen withdrawal in premature menopause is cognitively costly, and that replacement until natural menopause age (~51) is protective in this group Rocca 2007.

Mortality. Bayesian meta-analysis of women starting MHT <60 (n=8,400 across 19 trials): odds ratio 0.61 for all-cause mortality (95% CI 0.39–0.95) Salpeter 2009. Finnish national registry (n>500,000 women who used MHT): discontinuation associated with increased cardiac and stroke mortality, especially in users <60 stopping abruptly Mikkola 2015. WHI long-term: no all-cause mortality signal either direction at 18 years Manson 2017.

protocol

The 2022 NAMS position statement (the field's reference guideline) frames MHT as appropriate first-line therapy for symptomatic women under 60 or within 10 years of menopause, with shared decision-making documenting personal risk factors. Standard systemic regimens: oral conjugated equine estrogen 0.3–0.625 mg/day, oral estradiol 0.5–1.0 mg/day, transdermal estradiol patch 25–100 μg twice weekly, or gel/spray equivalents NAMS 2022 Stuenkel 2015. Transdermal route is preferred when VTE risk is non-trivial (BMI >30, prior VTE in family, age 60+), migraine with aura, hypertriglyceridemia, or active gallbladder disease Canonico 2007. Women with intact uterus require a progestogen: micronized progesterone 100 mg nightly (continuous-combined) or 200 mg for 12 days each month (sequential, often preferred in perimenopause). The levonorgestrel-releasing IUD is an off-label but increasingly accepted alternative for endometrial protection. Vaginal estrogen for GSM uses doses 5–10% of systemic with serum levels at or below postmenopausal baseline; can be combined with systemic MHT or used alone NAMS GSM 2020. Duration: NAMS 2022 explicitly rejects arbitrary stopping at 5 years — continuation should reflect ongoing risk-benefit, with no fixed ceiling; many symptomatic women benefit through their 60s and into their 70s if started early.

contraindications

Absolute contraindications to systemic MHT (NAMS, Endocrine Society, ACOG consistent): personal history of breast cancer, history of estrogen-sensitive cancer, history of venous thromboembolism or pulmonary embolism, history of stroke or myocardial infarction, active liver disease, unexplained vaginal bleeding, known or suspected pregnancy NAMS 2022 Stuenkel 2015. Relative contraindications: high baseline cardiovascular risk (consider transdermal), uncontrolled hypertension, gallbladder disease, hypertriglyceridemia, migraine with aura. Vaginal estrogen has a markedly narrower contraindication profile — the NAMS GSM statement notes safety even in breast cancer survivors after discussion with oncologist NAMS GSM 2020. The FDA 2025 announcement of boxed-warning reconsideration explicitly recognizes the class-extension problem: the warnings derived from WHI's older-mean-age systemic-CEE/MPA cohort have been applied to low-dose vaginal estrogen with no supporting evidence FDA 2025.

misconceptions

"HRT causes breast cancer." Combined E+P modestly increases risk (~1 extra case per 1000 women per year). Estrogen-alone in hysterectomized women decreases breast cancer incidence and mortality Anderson 2004 Chlebowski 2020. The distinction is regularly elided in public messaging.
"HRT causes heart attacks." Initiated within 10 years of menopause, MHT reduces coronary events and all-cause mortality. Initiated 20 years out in women with existing atherosclerosis, it raises early-period event risk Rossouw 2007 Hodis 2016 Boardman 2015.
"HRT has to stop at 5 years." NAMS 2022 explicitly rejects this — duration is individualized to risk-benefit, with no fixed limit NAMS 2022.
"Bioidentical compounded hormones are safer than FDA-approved MHT." No RCT evidence; compounded preparations lack dose verification and have been associated with endometrial hyperplasia in surveillance reports NAMS 2022.
"Vaginal estrogen carries the same risks as systemic." Serum levels with low-dose vaginal estrogen sit at or below postmenopausal baseline; the boxed warning class-extended to these products has no supporting trial data NAMS GSM 2020 FDA 2025.
"Hot flashes are just hot flashes — wait them out." Median duration is 7.4 years (the SWAN study); the women hardest hit (African American, early-onset symptom profile) average over 10 years NAMS 2022. Untreated vasomotor symptoms also correlate with cardiovascular endpoints in observational data, though causality is unclear.

audience

Population narrowing: all women in or past the menopausal transition. The strongest case for systemic MHT: women 40–59 within 10 years of menopause with bothersome vasomotor symptoms, GSM, or established osteopenia, no contraindications. Premature menopause (natural or surgical <45): systemic MHT recommended until at least age 51 by all major guidelines — withdrawal in this window is cardiovascularly and cognitively costly Rocca 2007 NAMS 2022. Late menopause (60+ initiation) where symptoms persist: case-by-case, transdermal preferred, lowest effective dose. Hysterectomized women: estrogen-alone, no progestogen needed, with the most favourable risk-benefit ratio of any MHT regimen in WHI data Anderson 2004. GSM alone: any age, low-dose vaginal estrogen, no contraindications except breast cancer history (and even there, after discussion with oncologist) NAMS GSM 2020.

failure-modes

Late initiation in symptomatic but high-risk women. The single biggest failure mode in the post-WHI era: women suffered through 5–10 years of vasomotor symptoms, then asked for MHT at age 65+ with established atherosclerosis, where the risk-benefit had genuinely flipped Manson 2024.
Wrong progestogen or no progestogen in a uterus-intact woman. Endometrial hyperplasia and cancer risk; the reason the FDA mandated combined regimens.
Compounded "bioidentical" preparations. Dose unverifiable; surveillance evidence of endometrial complications NAMS 2022.
Wrong route in high-VTE-risk women. Oral estrogen in a BMI 35 woman with prior superficial thrombophlebitis — transdermal is the correct choice and is often missed Canonico 2007.
Abrupt discontinuation. Vasomotor rebound is common and severe; the Finnish registry data also flags excess cardiac mortality in the first year after abrupt withdrawal in women <60 Mikkola 2015.
Under-dosing for symptom control. Standard-dose estradiol 0.5 mg often inadequate; titration to 1 mg or higher is appropriate when symptoms persist.

practicalities

Cost: generic oral estradiol and transdermal patch run roughly $10–40/month with insurance coverage in the US; brand-name combinations and bioidentical creams can run $100–300/month. Vaginal estrogen tablets and rings cost $50–250/month at list price, lower with insurance. Requires a prescriber — typically gynecologist, primary care physician, or menopause-certified clinician (NAMS Certified Menopause Practitioner directory). Initial visit, baseline mammogram and pelvic exam, optional baseline bone density. Follow-up at 3 months for dose titration, then annually. The structural friction is access: many primary care physicians trained during the 2002–2017 WHI-fear era are still reluctant to initiate MHT, and a menopause-fluent prescriber may take referral or self-pay Manson 2024.

history

Conjugated equine estrogen (Premarin) was approved in 1942 for vasomotor symptoms. By the 1990s, systemic MHT was the most prescribed drug class in postmenopausal American women, on the basis of observational data suggesting cardiovascular protection. The WHI (launched 1991, E+P arm halted 2002, E-alone arm halted 2004) was the first large RCT and reported the headline risks that triggered the 2003 FDA boxed warning Rossouw 2002 Anderson 2004. US prescriptions fell ~80% within two years. The 2007 age-stratified WHI re-analysis introduced the "timing hypothesis" Rossouw 2007; DOPS (2012), KEEPS (2014), ELITE (2016), and Manson's 18-year mortality follow-up (2017) progressively rehabilitated MHT for the early-postmenopausal population Schierbeck 2012 Harman 2014 Hodis 2016 Manson 2017. NAMS 2022 marked the formal pivot of the US menopause field NAMS 2022. The 2025 FDA expert panel and announced boxed-warning reconsideration is the regulatory acknowledgement that the warning, derived from a population with a median age of 63, was inappropriately class-extended FDA 2025 Manson 2024.

stakes

For the symptomatic 51-year-old who does not get MHT: a median 7.4 years of multiple-daily hot flashes and night sweats, fractured sleep, mood disruption, work-performance disruption, partner-relationship strain — the SWAN cohort confirms this is the modal experience NAMS 2022. GSM is progressive and does not remit: vaginal dryness, dyspareunia, recurrent UTI compound through the 60s and 70s. Bone loss during the menopausal transition is irreversible without intervention; women who lose 15% of femoral neck BMD in the first decade are on a different fracture trajectory at 80. The cardiovascular and dementia trajectories that an unreplaced premature/surgical menopause produces are documented and substantial Rocca 2007.

payoff

Vasomotor symptom relief is felt within 2–4 weeks of starting an adequate dose; sleep stabilizes by week 4–6; GSM (especially with vaginal estrogen) reverses by 4–12 weeks. Bone preservation is a multi-year accrual — net positive BMD trajectory measurable at 12–24 months. Cardiovascular benefit (in the early-initiator population) accumulates over years and shows up in the 10-year follow-up endpoints of DOPS and the 18-year Manson follow-up Schierbeck 2012 Manson 2017. Cognitive trajectory protection in surgical/premature menopause is multi-decade Rocca 2007. Mood and quality-of-life lift tracks vasomotor relief; the SSRI-class non-hormonal alternatives do not produce comparable global quality-of-life gains MacLennan 2004.

out-of-scope

Non-hormonal vasomotor therapies (SSRI/SNRI, gabapentin, oxybutynin, fezolinetant) — separate substances, separate entries. Testosterone supplementation for hypoactive sexual desire disorder in postmenopausal women — distinct intervention with thin evidence outside the libido endpoint. DHEA, tibolone (not available in US), and selective estrogen receptor modulators (raloxifene, ospemifene) — adjacent but separate. Lifestyle measures (cooling, layered clothing, weight management) — relevant adjunct but not what this entry covers. Bone-specific therapies (bisphosphonates, denosumab) — separate. Cognitive aging and dementia prevention generally — relevant only insofar as MHT timing modulates it.

The credibility range

Optimist case

For the typical reader — a woman 45–60 in or past the menopausal transition, symptomatic, without contraindications — MHT is one of the highest-value interventions in medicine: dominant relief of the dominant symptom (~75% hot-flash reduction), prevention of an osteoporotic fracture trajectory, reversal of GSM, plausible cardiovascular benefit when started early, plausible cognitive-trajectory benefit in premature menopause, plausible all-cause mortality benefit in the <60-at-initiation cohort. The WHI's headline risk numbers were derived from a population (mean age 63, median 12 years past menopause) that does not match the population now being treated; the post-2007 timing-stratified data is consistent across DOPS, ELITE, KEEPS, and the Cochrane stratified meta-analysis. The breast cancer signal is real but small in absolute terms (~1 case per 1000 women per year for combined therapy) and is reduced, not increased, by estrogen-alone in hysterectomized women. The boxed warning has produced two decades of preventable suffering — millions of women under-treated for vasomotor symptoms, accumulated bone loss, ongoing GSM, partner-relationship and career impact — for a population in which the risk-benefit was favourable. The 2025 FDA action acknowledges the error FDA 2025 Manson 2024.

Skeptic case

The timing-hypothesis evidence is observational re-analysis plus modest-sized RCTs (DOPS n=1006, ELITE n=643) that were not designed for hard cardiovascular endpoints. WHI itself, even age-stratified, did not show a statistically significant CHD reduction in the early-postmenopausal group — the favourable estimate is suggestive, not proven. The breast cancer increase with E+P is small but real and accumulates with duration; the CGHFBC meta-analysis showed risk continuing to elevate through 10+ years of use and persisting after stopping. WHI long-term follow-up: no all-cause mortality benefit either direction at 18 years Manson 2017. The "timing hypothesis" risks becoming a permission structure for indefinite MHT in healthy women who would otherwise pass through menopause with manageable symptoms; observational confounding (healthy-user bias, socioeconomic gradient) probably inflates the apparent benefit in early-postmenopausal cohorts. Commercial incentives are non-trivial: a large pharmaceutical market, a menopause-clinic industry, multiple consumer-facing brands. Vaginal estrogen for GSM is solidly evidenced and uncontroversial in current debate; the systemic-MHT enthusiasm has run ahead of trial data for hard endpoints.

Author's call

Land on the optimist side, with strong calibration. For symptomatic women under 60 or within 10 years of menopause without contraindications, MHT is appropriate first-line therapy and the field's two-decade reluctance was a regulatory error compounded by clinician training. The 2025 FDA action and NAMS 2022 position represent the genuine recalibration. For asymptomatic women, the case is weaker — bone-preservation alone does not earn an indefinite hormone prescription if the woman can do resistance training and adequate calcium/vitamin D. For women past 60 or 10 years out, the timing case has flipped and initiation is generally inappropriate. For vaginal estrogen specifically (GSM), the entry's recommendation is strong and uncontested in the literature; the FDA boxed warning on local products was never supported by trial data. Evidence rating: 4 — strong RCT base on symptoms, bone, and breast (WHI scale and quality is hard to surpass); cardiovascular and mortality remain partially-RCT, partially-observational. Controversy rating: 3 — the field has visibly recalibrated and major guidelines align, but the historical fear has cultural inertia and the timing hypothesis has measurable skeptics; vaginal-estrogen-vs-systemic is sometimes elided in lay discussion.

Stakeholder and incentive map

Menopause-medicine specialists (NAMS / The Menopause Society, IMS). Aligned on the timing-hypothesis recalibration; clinical and academic incentive to expand prescribing and de-stigmatize. Trial base they cite is genuine but the position is unmistakably reformist.
Pharmaceutical industry. Significant commercial interest in MHT (estradiol, vaginal estrogen rings/tablets, combination products, novel non-hormonal vasomotor drugs like fezolinetant). Industry-funded education has rebuilt clinician engagement post-WHI.
Primary care. Largely undertrained on MHT post-2002. Many PCPs still default to "the boxed warning is the boxed warning"; menopause-fluent care is currently a specialty referral problem.
Oncology and breast-cancer-survivor community. Conservative; the persistent E+P signal in WHI and CGHFBC keeps the breast cancer concern alive. The estrogen-alone reduction is not yet widely internalized in lay messaging.
Cardiovascular community. Recalibrating toward the timing hypothesis but historically wary; pharmacovigilance about VTE and stroke remains active.
FDA / regulatory. The 2025 action is a notable institutional pivot — the agency rarely re-opens a boxed warning. The reconsideration is targeted at the inappropriate class-extension to vaginal estrogen and the over-broad systemic warnings derived from older-cohort data FDA 2025.
Lay menopause-advocacy movement. A genuine consumer-led push (books, podcasts, social media) demanding access to MHT; sometimes oversells the case (e.g., framing MHT as universal prevention rather than symptom-driven treatment).
Skeptic counter-incentive. Academic statisticians and some epidemiologists urging caution on the timing hypothesis given the observational nature of the strongest re-analyses; the credibility-range skeptic case above is theirs.

Population variability

Time since menopause. The single biggest moderator. <10 years: favourable risk-benefit. >10 years or age 60+: shifts unfavourable, especially for cardiovascular and dementia outcomes Rossouw 2007 Hodis 2016.
Uterus status. Hysterectomized women receive estrogen alone, and have the most favourable WHI outcomes (including reduced breast cancer mortality) Anderson 2004.
Premature/surgical menopause. Stronger case for MHT through age 51 — withholding is a different harm profile from non-treatment of natural menopause Rocca 2007.
BMI / VTE risk. Transdermal route preferred; oral estrogen raises VTE risk additively with obesity Canonico 2007.
Race/ethnicity. SWAN data shows African American women have longer-duration vasomotor symptoms (median ~10 years); MHT access has historically been more limited in this subgroup — a real equity gap.
Personal/family breast cancer risk. High-risk women may still use vaginal estrogen for GSM; systemic MHT generally avoided in BRCA carriers post-prophylactic salpingo-oophorectomy except under specialist guidance.
Cardiovascular risk profile. Higher baseline ASCVD risk → transdermal preferred, threshold for initiation higher.

Knowledge gaps

Hard cardiovascular endpoint RCT in the early-postmenopausal population. The timing hypothesis rests on age-stratified WHI re-analysis plus modest-sized RCTs (DOPS, ELITE, KEEPS); a definitive RCT with MI/stroke as primary endpoint has never been run.
Micronized progesterone vs MPA breast safety. Observational data suggests micronized progesterone is safer, but no RCT has compared head-to-head with cancer endpoints CGHFBC 2019.
Cognitive effect of long-duration early-initiated MHT. WHIMSY found 7-year neutrality; longer follow-up at midlife-initiation cohorts is pending Espeland 2013.
Levonorgestrel IUD for endometrial protection in MHT. Widely used off-label; trial data thin.
Optimal duration. NAMS 2022 explicitly rejects fixed limits, but the lifetime-use risk-benefit balance past age 70 is sparsely studied.
Breast cancer survivor use of vaginal estrogen. Likely safe, evidenced by observational data and physiological reasoning; no RCT.
FDA 2025 outcome. Whether the boxed warning is removed, modified, or retained — and the downstream prescribing effect — is in active flux as of this writing FDA 2025.

Scope vs brief. The brief named vasomotor symptoms, bone, genitourinary tissue, cardiovascular risk, breast cancer risk, and cognition. The article covers all six explicitly, with the timing hypothesis and the 2025 FDA boxed-warning reconsideration as the structural spine. No silent narrowing.

Hard call on the FDA 2025 citation. As of writing (mid-2026), the FDA expert panel and the public announcement of reconsideration are documented; the boxed warning has been the subject of recent regulatory motion targeted especially at the inappropriate class-extension to vaginal estrogen products. The article phrases this as "announced reconsideration" and "the institutional movement is the news" to avoid overstating finality. If by the time this entry ships the boxed warning has been formally modified or removed, the relevant sentences should be tightened to reflect the action taken.

Hard call on evidence rating (4 vs 5). Considered 5 — there is a vast trial base, and the WHI alone is one of the largest RCTs in modern medicine. Held at 4 because the timing-hypothesis cardiovascular and mortality story rests on age-stratified re-analysis plus modest-sized purpose-built RCTs (DOPS n=1006, ELITE n=643). A dedicated hard-CV-endpoint RCT in the early-postmenopausal woman has never been run; the field is currently using stratified subgroup signals to drive practice. Rating 5 would imply that question is closed; it is not.

Hard call on controversy (3). The field's major bodies have aligned (NAMS 2022, Endocrine Society, IMS), which would argue for 2. Held at 3 because primary-care and lay-public framing still trails the specialist consensus by a wide margin, and the timing-hypothesis evidence has measurable academic skeptics. The post-WHI cultural inertia is a real signal of unresolved disagreement.

Hard call on longevity (3). Considered 2 because the WHI 18-year follow-up showed no all-cause mortality signal either direction Manson 2017. Held at 3 because the proven hip-fracture reduction in primary prevention is a longevity-relevant outcome on its own, and the early-starter Bayesian meta-analysis and Danish DOPS results land favourably even if the WHI long-run average is neutral. The Rocca 2007 premature-menopause signal further argues 3 is the honest read.

Contraindications token mapping. The catalogue's closed-vocabulary contraindication tokens (cardiac-condition, blood-thinners, pregnancy) do not include "personal breast cancer history" or "history of venous thromboembolism" — the actual hard MHT contraindications. Mapped to the closest available tokens. The article's contraindications section names the real list (breast cancer history, VTE history, stroke, MI, active liver disease, unexplained bleeding) in prose, which is the surface a reader actually reads. Flagging that the closed-vocabulary list could usefully be extended with `estrogen-sensitive-cancer-history` and `venous-thromboembolism-history` for this and adjacent entries.

Separate-entry candidates. Fezolinetant and the SSRI/SNRI / gabapentin / oxybutynin non-hormonal vasomotor options are large enough to warrant their own entry. Testosterone supplementation for postmenopausal hypoactive sexual desire disorder is a distinct intervention with its own evidence base. Tibolone (not US-available) is a separate hormone with a different risk profile. Each of these is signposted in out-of-scope and would link out when the entries land.

Future-link candidates. Resistance training (bone + sarcopenia overlap), vitamin D + calcium (bone), ApoB / lipid panel (the cardiovascular risk-stratification that informs the oral-vs-patch decision), pelvic floor physiotherapy. The related field currently lists only creatine as the most-existing adjacent musculoskeletal entry; the rest are placeholders for when those entries exist.

Audience scoping. Set gender: female, ages: ["40-59", "60+"]. The 18-39 band intentionally excluded; premature menopause cases in that age range are covered in the article as a subgroup but the substance does not apply to the typical 18-39 reader.

What the article doesn't cover that a maximalist version might. Specific brand-name comparisons (Premarin vs Estrace vs branded patches), the testosterone-add story for libido, perimenopausal contraception decision-making, and full discussion of compounded "pellet" therapy. All deliberately left out — brand specifics drift, testosterone is its own entry, contraception is its own field, and pellets do not have meaningful trial support.