One study. One formulation. A generation left to "tough it out."

The Golden Age

For four decades, estrogen therapy was ordinary medicine. Conjugated equine estrogens were first approved in the United States in 1942, and by the 1990s combined estrogen–progestin therapy was widely prescribed — for hot flashes, for bone protection, and, based on observational data, for a presumed protection of the heart.

By the end of the century the practice was mainstream: in 1999–2000, roughly 22% of postmenopausal women in the U.S. were using hormone therapy.³ Observational studies had repeatedly suggested cardiovascular benefit, and at times enthusiasm ran ahead of the evidence. The open question by 2000 was simple: would a large randomized trial confirm that hormones protected the heart?

July 2002: What the WHI Actually Tested

That trial was the Women's Health Initiative (WHI). On July 9, 2002, its investigators stopped the estrogen-plus-progestin arm early, reporting that combined therapy modestly raised the risk of breast cancer, coronary events, stroke, and blood clots.¹ The finding was real. But its specifics were narrow — and they were lost in the translation to headlines.

• One formulation. Oral conjugated equine estrogens (0.625 mg) plus a synthetic progestin, medroxyprogesterone acetate (2.5 mg) — a single combined pill.¹
• One population. 16,608 women aged 50–79, with an average age around 63 — for most, more than a decade past menopause.¹
• Small absolute risks. Per 10,000 women per year: 7 more coronary events, 8 more strokes, 8 more clots, and 8 more invasive breast cancers — set against 6 fewer colorectal cancers and 5 fewer hip fractures.¹

That last point is the one the coverage flattened. A relative risk increase of "26% for breast cancer" sounds alarming; the same finding expressed as fewer than one additional case per 1,000 women per year reads very differently. Both describe the identical result.

A separate WHI arm tested estrogen alone in women who had had a hysterectomy. It continued until 2004 and told a notably different story — no increase, and over longer follow-up a possible reduction, in breast cancer.² That nuance rarely reached the public.

The Great Retreat

The response was immediate and indiscriminate. Within nine months of the report, hormone-therapy prescriptions fell about 32%, and direct-to-consumer advertising essentially collapsed.⁴ The decline then held for a decade.

Crucially, the retreat extended even to therapies the WHI had never studied — estrogen-only regimens, transdermal patches, low-dose vaginal estrogen.⁵ A result about one pill in one population became a verdict on an entire class of medicine.

The deeper casualty was knowledge. As prescribing was discouraged, menopause faded from medical education, and a generation of clinicians finished training with little grounding in how to manage it. Europe, by contrast, responded more cautiously and continued to individualize therapy. The practical result in the U.S.: millions of women whose treatable symptoms were quietly reclassified as the ordinary price of getting older.

The Science Strikes Back

As the dust settled, researchers returned to the data — and the picture sharpened around two variables the original coverage had ignored: when therapy is started, and what is used.

Timing. The ELITE trial randomized healthy postmenopausal women to oral estradiol or placebo, grouped by time since menopause. Estradiol slowed the progression of early, subclinical atherosclerosis (measured as carotid-artery wall thickness) when started within six years of menopause — but not when started ten or more years out. Notably, it did not change CT measures of coronary calcium in either group.⁶ The "timing hypothesis" had its cleanest test: this is evidence on an early marker of disease, not yet on heart attacks themselves.

Formulation. The French E3N cohort found that estrogen combined with a synthetic progestin carried a higher breast-cancer risk (relative risk ~1.4) than estrogen combined with micronized (bioidentical) progesterone, whose risk was not significantly raised (~0.9).⁷ In other words, not all "HRT" behaves the same way — the progestogen matters.

Mortality. The most reassuring finding came from following the WHI participants for 18 years. Across roughly 27,000 women, hormone therapy was associated with neither an increase nor a decrease in overall mortality — 27.1% in the therapy groups versus 27.6% on placebo.⁸ Extended analyses of the trials refined the risk picture further by age and by specific outcome.⁹

The Quiet Reversal — and the 2025 Turn

For years the correction was quiet: guidelines shifted while the headlines did not. In 2022, the North American Menopause Society concluded that for women under 60, or within 10 years of menopause onset, and without contraindications, the benefit–risk balance favors treating bothersome symptoms and preventing bone loss; for those starting later, the balance is less favorable.¹⁰ In 2024, the WHI investigators themselves published a clinical-practice review consolidating the modern, age-stratified view.¹¹

Then, in 2025, the correction finally got loud. After a July 2025 advisory committee, the FDA announced on November 10, 2025 that it would remove the boxed ("black box") warnings concerning cardiovascular disease, breast cancer, and probable dementia from estrogen-containing menopausal hormone products.¹² It is not a blanket erasure: the agency retained the boxed warning for endometrial cancer on systemic estrogen-alone products, dropped the longstanding "lowest dose for the shortest duration" instruction, and added age-specific guidance reflecting the timing evidence.¹² Revised labels were expected to follow over the subsequent months.

What remains true is the part the whole saga obscured: hormone therapy carries real, individual risks — ones that depend on age, timing, formulation, route, and personal history — and it is also still the most effective treatment available for vasomotor symptoms and for the genitourinary syndrome of menopause.¹⁰ The lesson of the WHI was never "hormones are dangerous," nor "hormones are safe." It was that the nuance was the whole point.