In July 2002, a headline crossed every major news outlet: 'Hormone therapy causes breast cancer.' Within months, HRT prescriptions dropped by half. Over the following decade, an estimated 50,000 additional women in the US died from cardiovascular disease that hormone therapy might have prevented. The headline was wrong - or more precisely, it was a radical oversimplification of a study that never asked the question most women needed answered.
The Women's Health Initiative (WHI) trial remains the most consequential women's health study of the past 50 years - both for what it found and for how catastrophically the findings were misapplied. A 2023 re-analysis by the original investigators has clarified the picture significantly. Here is what it actually shows.
What the WHI actually studied
The WHI enrolled women averaging 63 years old - more than a decade past menopause - and randomised them to conjugated equine estrogen plus medroxyprogesterone acetate (a synthetic progestin), or placebo. The breast cancer finding that halted the trial was an additional 8 cases per 10,000 women per year. For context: that is approximately the same elevated risk as drinking one to two glasses of wine per day. It was statistically significant. It was not clinically catastrophic.
More importantly, the WHI was never a study of perimenopausal women deciding whether to start HRT. It was a study of women who were, on average, 63, postmenopausal for over a decade, and at elevated baseline cardiovascular risk. Applying those findings to a 48-year-old in early perimenopause is a category error.
The timing hypothesis - what the headline missed
The most important finding to emerge from the decade of WHI follow-up analyses is the 'timing hypothesis' or 'window of opportunity': hormone therapy started within 10 years of the final menstrual period, or before age 60, produces a substantially different risk-benefit profile than therapy started later.
The 2016 ELITE trial (Early versus Late Intervention Trial with Estradiol) made this concrete. Women who started oral estradiol within 6 years of menopause showed a 52% reduction in carotid intima-media thickness progression - a direct measure of atherosclerosis - compared to placebo. Women who started more than 10 years after menopause showed no benefit. Same drug, same dose, completely different outcome based solely on when treatment began.
What the 2023 re-analysis found
The 2023 re-analysis, published in JAMA and led by the original WHI investigators including JoAnn Manson, revisited the data with two decades of follow-up. Key findings for women aged 50–59 (the perimenopausal cohort within the WHI): estrogen-only therapy (in women without a uterus) was associated with significantly reduced breast cancer incidence - not increased. The combined estrogen-progestin arm showed a small increased breast cancer risk, concentrated in women who had used combination therapy before the trial.
The re-analysis also confirmed that younger, recently menopausal women showed net cardiovascular benefit - fewer deaths from heart disease and all-cause mortality. The risks that drove the 2002 headlines were concentrated in older women initiating therapy late.
What the progestin matters
The WHI used medroxyprogesterone acetate - a synthetic progestin with different receptor binding than natural progesterone. The CGHFBC meta-analysis and the French E3N cohort study both suggest that micronised progesterone (bioidentical) carries a lower breast cancer risk than synthetic progestins. This distinction is rarely mentioned in the popular-press coverage of HRT, but it is clinically meaningful and increasingly reflected in prescribing practice in Europe.
What current guidelines say
The North American Menopause Society 2022 position statement is explicit: hormone therapy is appropriate for symptomatic women who are within 10 years of menopause or under 60, without contraindications. The primary contraindication is personal history of hormone-receptor-positive breast cancer. The benefit-risk ratio is favourable for most women in this window, particularly for those with significant vasomotor symptoms or elevated cardiovascular risk.
Two decades of evidence have not overturned HRT - they have refined it. Timing matters. The type of progestogen matters. Delivery route matters (transdermal estrogen avoids the first-pass hepatic effects associated with oral preparations and may carry lower clot risk). The conversation your clinician should be having with you is nuanced, not binary.