The Women's Health Initiative

Contents

1. The instrument and its design
2. The 2002 results
3. The 2003 label
4. Reading the results by age
5. The 2025 relabeling
6. The range of normal

Menopause related concerns have run through the gynecologic work of Candace Cooley since she completed residency in 1998, the same decades in which the Women’s Health Initiative, a federal research program, reshaped how those concerns are treated. The program’s best known instrument was a pair of randomized hormone trials. In a randomized trial, chance rather than choice decides who takes the treatment and who takes a placebo, so differences in outcome can be credited to the treatment itself. That design is what gave the 2002 results their weight, and it is also why their limits took years to appreciate.

The instrument and its design

The program began in the early 1990s. A 2021 review in PubMed Central explains that its trials set out to test whether hormones prevent heart disease, with coronary heart disease as the main outcome, and that they enrolled women aged 50 to 79, with an average age of 63.¹ Women who still had a uterus received estrogen with a progestin or a placebo; women who had had a hysterectomy received estrogen alone or a placebo. The trials’ lead investigator told NPR in 2013 that recruiting in the 1990s was hard because about 40 percent of menopausal women were already taking hormones, and the final report followed 27,347 postmenopausal women through the two trials and the years after.² The Associated Press notes that in the 1990s more than 1 in 4 women in the United States took estrogen, often on the assumption that it would lower rates of heart disease and dementia.³

The broader program reached well beyond hormones. CNN reported in 2025 that it has produced more than 2,400 scientific publications, that its hormone findings are credited with preventing an estimated 126,000 breast cancer diagnoses, and that a women’s health researcher described its 1992 launch as a turning point away from treating women as small men in the study of chronic disease.⁴

Figure

The hormone trials over two decades

1. 01
   1993The lead investigator joins the study at its start
2. 02
   2002Estrogen plus progestin found to raise the risk of heart disease and breast cancer
3. 03
   2004Estrogen alone found to raise the risk of stroke, dementia and other problems
4. 04
   2013A final report breaks the results down by age and time since menopause

The 2002 results

The first results reversed the expectation. According to the 2013 NPR interview, the estrogen plus progestin trial reported in 2002 that the combination raised the risk of heart disease and breast cancer, and within months the number of women using hormones fell by almost half. A second report two years later found that estrogen alone raised the risk of stroke, dementia and other problems.² The review adds the hazard ratios from the combined arm: 1.29 for coronary heart disease and 1.41 for stroke.¹

The decline in use lasted. CNN reported in 2025 that prescriptions fell by more than 70 percent after the study and the label change that followed it, and that the share of postmenopausal women taking hormones dropped from more than 1 in 4 in the late 1990s to about 1 in 25 by 2020.⁵

The 2003 label

The trial’s results were written onto every product. STAT traces the boxed warning to a 2003 decision by the Food and Drug Administration, made after the 2002 study, which concentrated on combination therapy in older postmenopausal women but whose results were read broadly.⁶ CNN lists what the warning covered on pills, patches, sprays and creams alike: uterine and breast cancer, stroke, blood clots, and dementia in women over 65.⁵ The trials’ lead investigator later called that class wide labeling a mistake, since vaginal estrogen in low amounts does not raise blood levels.⁵ The same report notes that a 2022 review by the Preventive Services Task Force found no net benefit from either hormone regimen for preventing chronic conditions such as heart disease and dementia, and that four medical societies support hormones for symptoms but not for long term prevention.⁵

Figure

Postmenopausal women taking hormones

• Late 1990s, more than 1 in 425
• 2020, about 1 in 254

Reading the results by age

The second life of the trials came from dividing the results by age and by time since menopause. In the 2013 interview the lead investigator presents that breakdown as the key contribution of the final report: hormones should not be used to prevent chronic disease, she says, but younger women with moderate to severe symptoms can reasonably consider them, since they are less likely to have heart attacks, clots or strokes. The trials also showed that estrogen alone lowered breast cancer risk while the combination raised it, a difference she says is still not well explained.²

The 2021 review sets out the later numbers. Excess coronary risk fell mainly on women who began hormones after 60 or more than ten years after their final period; women in their 50s taking estrogen alone trended toward lower risk; and after 18 years of follow up there was no difference in cardiovascular or overall mortality between hormones and placebo. The review lists reasons the early results looked as they did: an average age of 63, at least twelve years on average since menopause, many women with existing risk factors, higher doses than later studies used, and many participants without hot flashes. Other outcomes shifted with time as well. Blood clots in the lungs and legs were clearly more common on the combination during treatment, with hazard ratios of 1.98 and 1.87, but over 13 years of follow up the difference from placebo was no longer significant. Over 18 years, stroke deaths did not differ. Both hormone arms also showed a lower risk of type 2 diabetes than placebo.¹ The Associated Press adds a September 2025 analysis finding no increase in heart problems for women in their 50s taking estrogen based drugs, a clear increase in their 70s, and an unclear picture in their 60s.³ The cardiac side is followed further in heart disease risk around menopause.

Figure

The two hormone trials

	Estrogen plus progestin	Estrogen alone
Results first reported	2002	2004
Headline risks	Heart disease and breast cancer	Stroke, dementia and other problems
Breast cancer	Risk increased	Risk reduced

The 2025 relabeling

In November 2025 the warning was removed. The Associated Press reports that the boxed warning came off more than 20 pills, patches and creams, that the original language moved lower on the label, and that a boxed warning remains telling women who still have a uterus to take a progestin with estrogen because estrogen given alone can lead to endometrial cancer.³ NBC News reports that labels for systemic therapy now recommend starting before 60 or within a decade of the onset of menopause.⁷

The change was debated. NPR reported that the agency received nearly 3,000 public comments, and that the American College of Obstetricians and Gynecologists had asked for formal advisory committee meetings and for separate treatment of vaginal and systemic estrogen.⁸ STAT quotes the lead authors of the 2002 paper, who took no position on the label but noted that no large randomized trials had been done since their own, and asked that information rest on rigorous science rather than anecdote.⁶ The lead investigator, for her part, told the Associated Press that the evidence for broad health benefits is not as conclusive as some officials claimed, but that without a warning applied to every product, attention might return to how the actual findings differ by age and underlying health.³

The range of normal

What the trials leave behind is less a single verdict than a range. A woman in her early 50s with bothersome hot flashes and a woman in her 70s seeking protection from chronic disease sit at different points in that range, and the lead investigator says a clinician would read the same data very differently for each. She describes the history as a pendulum that swung from hormones for every woman to hormones for none, and has settled somewhere between. For long term use in any age group, she adds, the bar stays very high because of the risk of clots and stroke.² The treatment itself is described in menopause hormone therapy, and the years before the final period in perimenopause. Where any one woman falls in that range is something she and her own doctor work out together, using results the trials themselves broke down by decade.