The 2002 Women’s Health Initiative found real risks from hormone therapy. It also got flattened into a warning that was too broad. The trial scared a generation of women, and their doctors, off a treatment that’s still reasonable for many of them. Here’s what WHI actually found, and how the risks break down.
What did the WHI actually find?
The trial enrolled about 27,000 women, but the average age was 63, roughly 12 years past menopause. Only about a third were in their 50s, the age when women actually start hormones for symptoms.1 When the results were announced as relative risks, “26 percent more breast cancer,” without the absolute numbers next to them, the increases sounded terrifying. In absolute terms, and especially for younger women, they were small. For women aged 50 to 59 on estrogen plus progestin, the trial found about 12 extra major outcomes per 10,000 women per year in its combined tally of risks and benefits, not 12 extra breast cancers; for women in their 70s, 38. And for women on estrogen alone, those without a uterus, the 50-to-59 group actually had 19 fewer major outcomes per 10,000 per year than placebo.2
How big were the risks?
Let’s be specific, because the absolute numbers are what matter. With estrogen plus progestin (the WHI used an oral conjugated estrogen with the synthetic progestin MPA), breast cancer rose by about 9 extra cases per 10,000 women per year, showing up after roughly 5 years of use.3 That’s about one extra case per 1,100 women each year, in the same range as the risk from drinking one to two glasses of wine a day. Blood clots ran about one extra per 475 women a year on the oral combined therapy, and stroke about one per 1,100.2 Those risks are real, but they’re small per person, and they concentrate in older women and in oral, synthetic-progestin regimens. On the other side of the ledger, hormone therapy cut hip fractures by about a third while women were taking it.5
Absolute-risk table: the WHI numbers in plain English
| WHI finding | Absolute framing | Why it matters |
|---|---|---|
| Combined estrogen plus progestin, age 50 to 59 | About 12 extra major outcomes per 10,000 women per year | Small absolute excess in the group closest to typical symptom treatment |
| Combined estrogen plus progestin, women in their 70s | About 38 extra major outcomes per 10,000 women per year | Age and time since menopause change the risk story |
| Estrogen alone, age 50 to 59 | About 19 fewer major outcomes per 10,000 women per year | Estrogen alone behaved differently from combined therapy |
| Breast cancer with combined therapy | About 9 extra cases per 10,000 women per year | Real risk, but smaller than the headline sounded |
| Hip fracture | About one-third lower while taking therapy | Bone protection was one of the clearer benefits |
Estrogen alone is a different drug, and so is the route
Two distinctions got lost in the coverage. First, estrogen-only therapy didn’t raise breast cancer in the WHI; over 20 years of follow-up it actually lowered both breast cancer cases and deaths (JAMA, 2020).3 It was the added synthetic progestin, MPA, that drove most of the breast cancer signal.4 Second, the route changes the risk. The WHI used oral pills, which pass through the liver and ramp up clotting factors. A patch or gel (transdermal estradiol) skips that step, and in observational data carries no increased clot or stroke risk at standard doses, versus a clear increase with the pills.67 Micronized progesterone, the body-identical form, also looks easier on clot risk than the older synthetic.6 None of the route and formulation data come from a fresh randomized trial, so it’s strong suggestion rather than proof, but the biology and the observations line up.
The limits of the WHI evidence
The trial still leaves a few limits. The idea that starting early protects your heart, the “timing hypothesis,” comes from re-slicing the old trial data after the fact, not a study built to test it; the USPSTF’s 2022 review calls it hypothesis-forming, not settled.8 The safety edge of patches and micronized progesterone is observational, because the big trial only tested oral pills with a synthetic progestin. The combined-therapy breast cancer risk doesn’t vanish the moment you stop; it lingers for years.4 And we have thin data on using hormones well past 7 years, or on starting them well after 60. Symptom relief and bone protection are proven. Earlier start, patches, and micronized progesterone are plausible risk improvements, but not proven by a new randomized trial.
Use hormone therapy for symptoms, with the safer setup
If you’re within about 10 years of menopause, under 60, and have symptoms worth treating, hormone therapy is a reasonable and often good choice, and the risks are smaller than you were probably told. Ask about a transdermal patch or gel rather than a pill, and micronized progesterone rather than an older synthetic if you still have a uterus; that combination has the most favorable risk profile we know of. Know the hard stops: a history of breast cancer, blood clots, a recent stroke or heart attack, or active liver disease mean systemic hormones aren’t for you. If you’re mainly deciding whether to start at all, I walk through the bigger picture in the perimenopause piece.