The WHI study explained: what it found and what changed since

The Women’s Health Initiative (WHI) was a set of US government-funded randomized trials that tested whether hormone therapy prevents heart disease and other chronic illness in postmenopausal women aged 50 to 79. In 2002 its estrogen-plus-progestin arm was stopped early because risks outweighed benefits, and hormone therapy use collapsed. Two decades of follow-up have since shown that the harms were concentrated in older women, that hormone therapy did not increase deaths, and that estrogen alone lowered breast cancer risk. Those findings reshaped guidelines and, in 2025, the FDA label.

What the WHI was

The National Institutes of Health enrolled 27,347 postmenopausal women at 40 US centers between 1993 and 1998. Their average age was 63.4. Women with a uterus (16,608) received conjugated equine estrogens 0.625 mg plus medroxyprogesterone acetate 2.5 mg daily, or placebo. Women who had had a hysterectomy (10,739) received the same estrogen alone, or placebo.

Three design facts explain much of what followed.

• It was a prevention trial. The primary outcome was coronary heart disease. Women were not recruited because they had hot flashes; many had none.
• The participants were older than typical users. With an average age of 63, they were on average more than a decade past menopause.
• It tested one regimen. Oral conjugated estrogen with a synthetic progestin was the standard US product in the 1990s. Estradiol patches and micronized progesterone, common today, were not studied.

What the 2002 report found

After a mean of 5.2 years, the data and safety monitoring board stopped the combined arm. Per 10,000 women per year, estrogen plus progestin was associated with:

Outcome	Difference vs. placebo	Hazard ratio
Coronary heart disease	+7	1.29
Stroke	+8	1.41
Pulmonary embolism	+8	2.13
Invasive breast cancer	+8	1.26
Colorectal cancer	−6	0.63
Hip fracture	−5	0.66
Death from any cause	No difference	0.98

The net “global index” was 19 excess events per 10,000 women per year. The authors concluded the regimen should not be started or continued to prevent heart disease. That conclusion was correct and still stands.

The 2004 estrogen-alone results

The estrogen-alone arm ran until 2004, an average of 6.8 years. Estrogen alone did not raise heart disease (hazard ratio 0.91), lowered hip fractures by 6 per 10,000 women per year, raised stroke by 12 per 10,000 per year, and showed a nonsignificant 23 percent reduction in breast cancer. The burden of disease was the same in both groups: no overall benefit for prevention, and no overall harm.

Timing: the analysis that changed the interpretation

In 2007 the investigators reanalyzed both trials by years since menopause. For coronary heart disease:

Years since menopause at start	Hazard ratio	Absolute difference per 10,000 per year
Fewer than 10	0.76	−6
10 to 19	1.10	+4
20 or more	1.28	+17

Stroke risk rose by about a third regardless of timing. For women aged 50 to 59, total mortality was lower on hormone therapy (hazard ratio 0.70), with a trend across age groups that did not reach formal significance. The authors wrote that these data “should be considered in regard to the short-term treatment of menopausal symptoms.” This timing hypothesis is now the basis for the guideline window of under 60 or within 10 years of menopause.

Long-term follow-up

Mortality. Through 2014, cumulative 18-year follow-up of all 27,347 women found that 27.1 percent had died in the hormone groups versus 27.6 percent on placebo, a hazard ratio of 0.99. Cardiovascular and cancer deaths were also unchanged. Five to seven years of hormone therapy neither shortened nor lengthened life.

Breast cancer. With more than 20 years of follow-up, the two regimens diverged:

	Estrogen plus progestin	Estrogen alone (after hysterectomy)
Breast cancer incidence per year	0.45% vs 0.36% placebo (HR 1.28)	0.30% vs 0.37% placebo (HR 0.78)
Breast cancer deaths	71 vs 53 (HR 1.35, not significant)	30 vs 46 (HR 0.60, significant)

In plain terms, the WHI regimen of combined therapy produced roughly 9 extra breast cancers per 10,000 women per year, while estrogen alone produced about 7 fewer. How that translates to an individual is covered in Hormone therapy and breast cancer risk.

What changed in practice

Use fell fast and stayed low. National survey data show 26.9 percent of US postmenopausal women used hormone therapy in 1999 to 2000 and 4.7 percent in 2017 to 2020, with the steepest drop among women aged 52 to 65. Among non-Hispanic Black women use fell from 11.9 percent to 0.5 percent, and among Hispanic women from 13.8 percent to 2.6 percent.

Guidelines split the question in two. For preventing chronic disease, the answer from the trials themselves stayed no. For treating bothersome symptoms in women under 60 or within 10 years of menopause, US professional societies concluded that the benefits outweigh the risks. In November 2025 the FDA began revising hormone therapy labels along the same lines; we cover that in Is hormone therapy safe?.

What the WHI could not tell us

• It did not study estradiol patches or gels, or micronized progesterone, so their long-term risks are inferred from observational data.
• It did not include perimenopausal women, and relatively few participants were in their early 50s with symptoms. See hormone therapy in perimenopause vs after menopause.
• It tested 5 to 7 years of use; longer durations rely on observational studies.
• A single trial, however large, cannot settle every question. Its results are strongest for the regimen and age group it tested.

Questions to ask

• Which hormones and route are you suggesting, and how do they differ from what the WHI tested?
• How far am I from my last period, and how does that change my numbers?
• Do my personal risks, such as a clotting or breast cancer history, change the picture?
• How will we weigh symptom relief against these small absolute risks for me specifically?

If you are starting this conversation, our symptom check produces a summary to bring, and our OB-GYN directory lists clinicians near you.