A single study in 2002 changed how a generation of women were treated, and it was badly misread.
What happened
The announcement reached the evening news before it reached the medical journals. The Women's Health Initiative was stopped early in 2002 when data showed an increase in breast cancer among women taking estrogen with a progestin. Not just any estrogen and progestin: the trial used a synthetic progestin, medroxyprogesterone acetate, paired with estrogen extracted from horse urine. The number announced was a 26 percent increase in breast cancer.
Two footnotes never made the broadcast. The number did not reach statistical significance, meaning that by the ordinary rules of evidence it could not be separated from chance. It was also a relative risk: in absolute terms, roughly 8 additional cases per 10,000 women per year, which is 0.08 percent.
What the press conference did not lead with: the women in the trial's other arm, who took the horse estrogen alone without the synthetic progestin, went on to show a decreased risk of breast cancer.
The damage was immediate. Within a single year, prescriptions for the exact drug studied fell by two thirds, and a generation of women walked away from hormones over a number that could not stand on its own.
Why it does not apply the way people think
The average woman studied was about 63 and many were ten to fifteen years past menopause. That is not the same person as a symptomatic woman in her early fifties considering therapy near the onset of menopause.
The drugs were different too. Oral conjugated equine estrogens with a synthetic progestin are not the bioidentical hormones we use today. Transdermal estradiol delivered through the skin avoids the first pass through the liver that raises clotting factors, and bioidentical micronized progesterone behaves differently from the progestin studied.
There is also a statistical problem buried in the trial that almost nobody mentions. About a quarter of the women had taken hormones earlier in life, and the prior users assigned to placebo went on to have fewer breast cancers than expected, fewer than the general population. An unusually healthy comparison group inflates the apparent risk of whatever you compare against it. Among the women who had never taken hormones before, the majority of the trial, the difference between the two groups was minimal and could not be distinguished from chance.
Women who took estrogen alone showed no increase in breast cancer, and among the younger women there was a reduction in all cause mortality.
Where it stands now
Everything the headline ignored turns out to matter: age at initiation, time since menopause, formulation and route. Started near the onset of menopause with bioidentical estradiol delivered through the skin and bioidentical progesterone, the balance of benefit and risk is favorable for most women.
Women who start within ten years of menopause show decreased cardiovascular risk, fewer urinary tract infections, less osteoporosis, and potentially a decreased risk of dementia, though studies there have pointed in different directions. The window does not slam shut at ten years, either. The newer data has opened that conversation to women many years past menopause, who can weigh their real risks against their real benefits with a provider who knows this literature and make a shared decision. Bone protection and UTI prevention hold up at older ages, and quality of life often improves too.
The list of true contraindications is shorter than most women have been told. The clotting concern that ruled out so many belongs to oral estrogen, and delivering estradiol through the skin largely solves it. A history of breast cancer narrows the menu without emptying it, since testosterone and other options often remain. Local vaginal estrogen can be used in all women, with no contraindications to speak of. Almost nobody is truly out of options.
If you are wondering whether your own symptoms fit this picture, my free symptom check walks through ten areas in about five minutes, scores each one, and gives you a page to print for your appointment.