Women's Health Initiative: Estrogen plus Progestin Hormone Therapy

US prescriptions for postmenopausal hormones rose from 58 million in 1995 to 90 million in 1999, about 15 million women a year, and most of the growth came from combined estrogen-progestin pills, chiefly Prempro. Its popularity rested largely on observational data, particularly the Nurses' Health Study, which suggested that postmenopausal women taking hormones had substantially lower rates of coronary disease. That signal was strong enough that preventing heart disease became a leading reason to prescribe, alongside relief of menopausal symptoms. The Women's Health Initiative was designed to test that assumption in a proper randomized trial.

The estrogen-plus-progestin arm enrolled 16,608 postmenopausal women with an intact uterus at 40 clinical centers across the United States, randomizing them to conjugated equine estrogen at 0.625 mg plus medroxyprogesterone acetate at 2.5 mg daily, or matching placebo. Planned follow-up was 8.5 years. Jacques Rossouw of the National Heart, Lung, and Blood Institute was the project officer and first author of the main report; Garnet Anderson coordinated the statistical analysis, and JoAnn Manson was principal investigator at the Brigham and Women's Hospital site.

On May 31, 2002, after a mean of 5.2 years, the data and safety monitoring board recommended stopping the estrogen-progestin arm because the test statistic for invasive breast cancer had crossed its stopping boundary; the hazard ratio for breast cancer was 1.26, or about eight additional cases per 10,000 person-years. Coronary heart disease, stroke, and pulmonary embolism were all elevated in the hormone arm, while hip fractures and colorectal cancers were less frequent. There was no cardiovascular benefit, and the results contradicted directly the observational literature that had driven prescribing for nearly two decades.

The JAMA paper appeared in July 2002 and drew wide attention outside medicine. Prescriptions for Prempro, the combined pill used in the trial, fell by 66% in the first half of 2003 compared with the first half of 2002, and Premarin prescriptions fell by 33%. Many patients contacted their physicians to stop treatment; many physicians stopped initiating it. Total US hormone prescriptions were projected to fall to about 57 million in 2003, back to the level of 1995.

Participants had a mean age of 63.3, and many had been postmenopausal for years before enrollment. A 2007 secondary analysis of both WHI hormone trials found a hazard ratio for coronary heart disease of 0.76 in women less than 10 years past menopause and 1.28 in women 20 or more years past it, the basis of the timing hypothesis. The North American Menopause Society's 2022 position statement judged the benefit-risk ratio favorable for treating symptoms in women under 60 or within 10 years of menopause, and less favorable for women who start later.