Perimenopause and Hormone Therapy: How Two Decades of Evidence Reshaped the Conversation

A patient in her late forties begins waking at 3 a.m. drenched in sweat. Her periods have become unpredictable — heavier some months, absent others. Her sleep is fragmenting, her mood has shifted in ways her family has noticed before she has, and her clarity at work has visibly dropped. She raises this with her primary care physician and is told that menopause is a natural transition, that hormone therapy carries cancer risk, and that she should try cooling sheets and meditation.

This conversation, repeated millions of times over the past twenty years, traces back to a single trial that was widely misinterpreted, applied to populations it was never designed to study, and left a generation of clinicians and patients with a risk-benefit picture that the underlying data does not actually support. The reassessment has been quietly underway for more than a decade. Most patients have not heard about it.

What Perimenopause Actually Is

Perimenopause is the menopausal transition itself — the period beginning with the first irregularities in the menstrual cycle and ending one year after the final menstrual period. It typically spans four to ten years and most often begins in the mid-forties, though earlier onset is common. Menopause itself is a single retrospective date: the twelve-month anniversary of the last period.

The hormonal picture during perimenopause is not a smooth decline. Estradiol fluctuates dramatically, sometimes reaching higher peaks than during regular cycling and dropping to postmenopausal levels within the same month. Progesterone falls earlier and more linearly as ovulatory cycles become rarer. The combination produces the symptom cluster most patients recognize — vasomotor symptoms (hot flashes and night sweats), sleep disruption, mood changes, brain fog, vaginal dryness, joint pain, irregular and often heavy bleeding, and a marked rise in migraine frequency in patients who are migraine-prone.

The North American Menopause Society 2022 position statement noted that roughly 75% of women experience vasomotor symptoms during the transition, and approximately one-third of those describe symptoms severe enough to interfere with daily functioning. Median duration of bothersome vasomotor symptoms in the SWAN cohort was 7.4 years. The framing of menopausal symptoms as a brief inconvenience does not match the evidence on duration.

The WHI Story, Properly Told

The 2002 Women's Health Initiative (WHI) trial is the single most consequential moment in modern women's health communication. The combined estrogen-progestin arm was halted early after interim analysis suggested elevated breast cancer and cardiovascular event risk. The findings were communicated to the public as evidence that hormone therapy was dangerous. Prescriptions collapsed by roughly 70% within two years and have only partially recovered.

The full picture is more complicated. The WHI enrolled women with a mean age of 63, more than a decade past the average age of menopause. The trial was designed to test whether hormone therapy initiated late in life prevented chronic disease, not whether it managed symptoms in newly menopausal women. The patients who would actually be candidates for therapy in clinical practice — symptomatic women in their late forties and early fifties — were a small minority of the trial population.

When Rossouw and colleagues (JAMA, 2007) reanalyzed the WHI data stratified by age and time since menopause, the picture changed substantially. Women who initiated combined hormone therapy within ten years of menopause showed no increase in coronary heart disease and a reduction in all-cause mortality compared with placebo. Women who initiated more than ten years after menopause showed the elevated cardiovascular risk that drove the original termination. This is now known as the timing hypothesis: the same therapy is beneficial in early menopause and potentially harmful in late initiation.

The ELITE trial (Hodis and colleagues, NEJM, 2016) tested the timing hypothesis directly by randomizing women either within six years of menopause or more than ten years past menopause to oral estradiol versus placebo. After five years, the early-initiation group showed significantly slower progression of carotid intima-media thickness, a validated marker of early atherosclerosis. The late-initiation group showed no such benefit. The biological explanation — that estrogen exerts vasculoprotective effects on healthy endothelium but can destabilize established atherosclerotic plaque — has been mechanistically supported by subsequent work.

What the Long-Term Mortality Data Actually Shows

In 2017, Manson and colleagues published the 18-year cumulative follow-up of the WHI in JAMA. The headline result, largely unreported in mainstream coverage, was that there was no increase in all-cause mortality in either the combined estrogen-progestin arm or the estrogen-alone arm compared with placebo. For women who started therapy aged 50 to 59 — the clinically relevant subgroup — all-cause mortality was numerically lower in the hormone therapy groups, though not statistically significant in either arm individually.

Mikkola and colleagues, working with Finnish national registry data on nearly half a million women using postmenopausal hormone therapy, found that women who used hormone therapy for any duration had lower breast cancer mortality and lower all-cause mortality than non-users when followed for at least ten years. Registry data carries different limitations than randomized trial data, but the consistency of the all-cause mortality signal across study designs is striking.

The reframed evidence base now supports a working consensus: in healthy women under 60 or within 10 years of menopause, the benefits of hormone therapy for symptomatic relief outweigh the risks for most patients. The North American Menopause Society 2022 statement makes this explicit and adds that the benefit-risk ratio remains favorable for continued use beyond age 60 in many women, though the calculus becomes more individual.

The Categories of Hormone Therapy

Modern menopausal hormone therapy is not a single intervention. The clinically important distinctions:

Estrogen alone is appropriate for women without a uterus. It is most often prescribed as transdermal estradiol (patch, gel, or spray) or oral estradiol. Transdermal delivery bypasses first-pass hepatic metabolism and is associated with lower venous thromboembolism risk than oral preparations — a difference that has become a guideline-level preference in patients with elevated baseline clotting risk.

Estrogen plus a progestogen is required for women with an intact uterus, because unopposed estrogen exposure stimulates endometrial proliferation and increases endometrial cancer risk. The choice of progestogen matters. Micronized progesterone, identical in structure to endogenous progesterone, is now the preferred option in most guidelines and appears to carry a more favorable breast and cardiovascular risk profile than the synthetic medroxyprogesterone acetate used in the original WHI trial.

Low-dose vaginal estrogen is a fundamentally different therapy. Used for genitourinary syndrome of menopause — vaginal dryness, painful intercourse, urinary symptoms — it produces minimal systemic absorption and is not subject to the same risk-benefit considerations as systemic therapy. It is appropriate for women with mild systemic symptoms who only need genitourinary relief, and increasingly considered safe even in many women with a history of hormone-receptor-positive breast cancer (in consultation with their oncology team).

The Symptoms That Are Actually Treatable

Hormone therapy is the most effective treatment available for moderate-to-severe vasomotor symptoms, with reductions in frequency and severity in the range of 75 to 90% in randomized trials. No non-hormonal therapy approaches that effect size. SSRIs, gabapentin, and the newer NK3 receptor antagonist fezolinetant produce meaningful but smaller reductions and are appropriate for women who cannot or prefer not to use hormones.

Sleep disruption driven by night sweats responds to hormone therapy. Sleep disruption with other drivers — primary insomnia, obstructive sleep apnea, depression — warrants its own evaluation.

Genitourinary symptoms respond reliably to vaginal estrogen and often improve substantially with systemic therapy as well. These symptoms tend to worsen rather than improve over time without treatment, which is one reason the conventional advice to "wait it out" produces particularly poor outcomes here.

Mood symptoms during the perimenopausal transition — when estradiol fluctuations are at their largest — show meaningful response to hormone therapy in patients without a primary depressive disorder. Bone density preservation is a documented effect of systemic therapy and was the original FDA indication for menopausal hormone use; the bone benefit accrues automatically when therapy is prescribed for symptom management.

Where the Risks Actually Sit

The risks are real and worth naming specifically.

Breast cancer. Combined estrogen-progestogen therapy is associated with a small absolute increase in breast cancer risk after roughly five years of use. The WHI estimated roughly 8 additional cases per 10,000 women per year of use. Estrogen-alone therapy has not shown the same signal and in some analyses has been associated with reduced breast cancer incidence. The progestogen choice matters, with micronized progesterone associated with a more favorable profile than medroxyprogesterone acetate.

Venous thromboembolism. Oral hormone therapy roughly doubles the risk of deep vein thrombosis and pulmonary embolism, though absolute risk in healthy women under 60 remains low. Transdermal preparations do not show this increase in most studies and are preferred in patients with elevated baseline thrombotic risk (obesity, prior clot, certain inherited thrombophilias).

Stroke. A small absolute increase in ischemic stroke is observed with oral therapy, particularly in older initiators. Transdermal preparations again appear to mitigate this signal.

Endometrial cancer. Risk is meaningfully elevated only when estrogen is given without adequate endometrial protection (a progestogen for women with a uterus). Modern combined regimens neutralize this risk.

The risks above are absolute numbers in specific populations, not abstract warnings. For a healthy 50-year-old without contraindications, the cumulative absolute risk of any of these events over five years of therapy is small. For a 70-year-old with uncontrolled hypertension, prior breast cancer, or active liver disease, the calculus shifts decisively.

The Conversation That Should Be Happening

A patient presenting with bothersome perimenopausal symptoms deserves a structured discussion that includes her actual symptom burden, her individual risk profile (cardiovascular, thrombotic, oncologic), the available preparations and routes, and the timing of her menopausal transition. The 2022 NAMS statement, the 2022 ACOG guidance, and the British Menopause Society recommendations are aligned on this approach.

What the evidence does not support is a blanket refusal to discuss hormone therapy because of a 23-year-old trial that studied a different population and used a different progestogen at a different starting age. Patients who are turned away from this conversation often spend years cycling through inadequate alternatives, and many ultimately reach menopause with worse sleep, worse cardiovascular trajectories, worse bone density, and worse genitourinary health than they would have otherwise.

The reassessment is not a recommendation that every menopausal woman should take hormones. It is a recommendation that the question deserves an actual answer, calibrated to the patient in front of you, rather than a reflexive no.

Dr. Anika Sharma is the Women's Health Editor at HealthKoLab. She is a board-certified obstetrician-gynecologist with a decade of clinical experience and particular interest in hormonal health across the female lifespan.