Where the fear came from
Meet the study everyone's afraid of: the WHI
In 1991, the U.S. government launched the Women's Health Initiative (WHI), one of the largest health studies of women ever done. Part of it tested whether hormone therapy helped prevent disease in postmenopausal women. Over 27,000 women took part.
In 2002, researchers stopped one part of the study early. Women taking a specific hormone combination, estrogen plus a synthetic progestin called medroxyprogesterone acetate (MPA), sold in combination with estrogen under the brand name Prempro, appeared to have a higher rate of breast cancer than women taking a placebo (a fake pill). Depending on which WHI publication you look at, the increase is reported as 24–26% higher relative to the placebo group.
That number hit the news everywhere. Almost overnight, "hormones cause breast cancer" became conventional wisdom. Millions of women stopped their prescriptions. Doctors stopped recommending hormone therapy. That single headline has shaped medical advice for more than twenty years.
The part that didn't make headlines
What the "26% increase" actually meant: and what it left out
A "26% increase" sounds alarming. But a few important details got lost on the way to the evening news.
- The women in the study were much older than typical hormone therapy patients. The average age of participants was 63, already more than a decade past when most women start hormone therapy, which is typically in their late 40s or early 50s, right around menopause. Only about a third of participants (32%) were even in their 50s; 45% were in their 60s, and 22% were in their 70s. That matters, because starting hormone therapy soon after menopause looks different, biologically, than starting it many years later.
- "26% higher" was a small number made to sound large. Relative risk, the "26% increase", describes a change compared to a small starting number, not a jump to a large one. In real terms, it worked out to roughly 8 extra cases of breast cancer per 10,000 women per year. That's a genuine, real signal worth taking seriously, but it's a very different picture than "hormones cause breast cancer," full stop.
- Only one hormone combination was tested. The study used exactly one recipe: estrogen plus medroxyprogesterone acetate, a synthetic progestin. It did not test the many other forms of estrogen and progesterone available today, including micronized progesterone, which is molecularly identical to what your own body makes, and which we now know behaves differently in breast tissue than the synthetic version used in the WHI.
- The estrogen-only group told a completely different story. Women in the WHI who'd already had a hysterectomy didn't need a progestin to protect their uterus, so they were given estrogen alone. That group did not have a higher rate of breast cancer. In fact, their risk was lower than the placebo group's. A 20-year follow-up of these women, published years later, confirmed both a lower rate of breast cancer diagnoses and lower breast cancer deaths compared to women who took no hormones at all. This finding rarely made the news, "hormones don't cause cancer, and may even help" isn't as catchy a headline.
Two decades of follow-up research
What newer, larger studies have found
Since 2002, researchers have followed up with much more detail, and a clearer picture has emerged.
The type of progesterone matters
A large French study that followed thousands of women for years (the E3N cohort) found that women taking estrogen with micronized progesterone, again, the form identical to your own hormone, had no measurable increase in breast cancer risk in the first five years, and only a modest increase after that. Women taking estrogen with a synthetic progestin (like the one used in the WHI) had a clearly higher risk. This is now one of the most consistent findings in the field: it's not "hormones" as a category that drive the risk, it's specifically the synthetic progestin pairing.
The newest and largest analysis to date
In early 2026, researchers published the most comprehensive review yet: a combined analysis of 34 studies covering more than 4.5 million women. It confirmed the same pattern. Estrogen paired with a progestin carried a real, modest increase in risk, but estrogen alone did not, and in randomized trials actually appeared to lower risk slightly. The analysis also found something reassuring for anyone reconsidering hormone therapy: women who had stopped using it showed no continued increase in risk. Whatever risk exists while using estrogen-progestin therapy appears to fade after stopping.
The FDA updated its guidance to reflect this
In February 2026, the FDA removed the "black box" breast cancer warning that had been on hormone therapy labels since 2003, the strongest type of safety warning the FDA issues. This wasn't a political decision; it reflects more than two decades of additional data showing the original blanket warning didn't capture how much risk actually depends on the specific hormone, the age it's started, and how long it's used.
What this looks like in real numbers
Breast cancer risk by age: 1,000 women, each decade
Every percentage above is a change on top of your starting point, which rises with age regardless of hormone use. Here's what that starting point looks like, and how it shifts with different types of hormone therapy.
In your 40s
~15cases per 1,000 women over 10 years985 out of 1,000 women in their 40s will not be diagnosed with breast cancer in the next decade, with or without hormone therapy.
No hormones
~15 out of 1,000
- Baseline cases (15 per 1,000)
- No diagnosis
Body-identical progesterone, under 5 yrs
~17 out of 1,000
- Baseline cases (15 per 1,000)
- Additional cases (2 per 1,000)
- No diagnosis
Synthetic progestin (WHI-type)
~22 out of 1,000
- Baseline cases (15 per 1,000)
- Additional cases (7 per 1,000)
- No diagnosis
In your 50s
~23cases per 1,000 women over 10 years977 out of 1,000 women in their 50s will not have a diagnosis in the next decade. This is the decade most hormone therapy research has focused on.
No hormones
~23 out of 1,000
- Baseline cases (23 per 1,000)
- No diagnosis
Body-identical progesterone, under 5 yrs
~26 out of 1,000
- Baseline cases (23 per 1,000)
- Additional cases (3 per 1,000)
- No diagnosis
Synthetic progestin (WHI-type)
~37 out of 1,000
- Baseline cases (23 per 1,000)
- Additional cases (14 per 1,000)
- No diagnosis
In your 60s
~35cases per 1,000 women over 10 yearsBackground risk keeps rising with age either way. This is roughly the age group the WHI studied, a good reason its results don't map cleanly onto women starting hormone therapy closer to menopause.
No hormones
~35 out of 1,000
- Baseline cases (35 per 1,000)
- No diagnosis
Body-identical progesterone, under 5 yrs
~40 out of 1,000
- Baseline cases (35 per 1,000)
- Additional cases (5 per 1,000)
- No diagnosis
Synthetic progestin (WHI-type)
~57 out of 1,000
- Baseline cases (35 per 1,000)
- Additional cases (22 per 1,000)
- No diagnosis
Starting-point risk: American Cancer Society, Breast Cancer Facts & Figures 2024–2025, and SEER population data. Hormone therapy figures: approximate, synthesized from the E3N cohort's published analyses (Fournier et al., 2005 and 2008; exact estimates vary somewhat by publication and duration cutoff, generally ranging from no significant increase to a modest one) for body-identical progesterone, and the 2026 meta-analysis of 34 studies / 4.5 million women, Annals of Medicine, for synthetic progestin combinations.
Putting the numbers in perspective
How hormone therapy compares to everyday risk factors
Numbers like "26% higher risk" are hard to judge in isolation. Here's the same kind of risk increase, side by side with other things that affect breast cancer risk, for a woman in her 50s.
Estrogen alone
Randomized trial data — lower risk, not higher
No hormones (starting point)
Reference
Body-identical progesterone, <5 yrs
Not a significant increase
Smoking
Evidence inconsistent — see note
2+ alcoholic drinks / day
Consistent, real increase
Body-identical progesterone, >5 yrs
Modest, real increase
Obesity
Consistent, real increase
Synthetic progestin (WHI-type)
Clearly elevated, currently using
Cases per 1,000 women in their 50s, over 10 years.
Obesity: Neuhouser et al., JAMA Oncology, 2015 (Women's Health Initiative data); Renehan et al., The Lancet, 2008. Alcohol: British Menopause Society / Women's Health Concern factsheet; Collaborative Group on Hormonal Factors in Breast Cancer, British Journal of Cancer, 2002 (53 combined studies). Smoking: Reeves et al., Generations Study, Breast Cancer Research, 2017. see note below. Hormone figures: as cited above.
Where this leaves you
The short version
| What the research shows | What it means |
|---|---|
| The WHI (2002) | One hormone combo, in older women, showed a modest increase, widely overgeneralized |
| Estrogen alone (WHI, 20-yr follow-up) | Lower breast cancer rates and deaths, not higher |
| Body-identical progesterone (E3N cohort) | No measurable increase under 5 years; modest after |
| Synthetic progestins (multiple studies) | Consistently higher risk than body-identical progesterone |
| 2026 meta-analysis (4.5M women) | Confirms the pattern above; risk fades after stopping |
| FDA (Feb 2026) | Removed the 2003 black-box breast cancer warning |
None of this means hormone therapy is risk-free, it isn't, and the type of progesterone, your age, and how long you use it all genuinely matter. But the blanket fear that took hold in 2002 was built on one hormone combination, tested in a much older population, and it left out the parts of the story that would have made the picture far less frightening. Twenty years of additional research, and the FDA's own updated guidance, now reflect that.
This is population-level information. Your own history, family history, and goals are what should shape your personal decision, in conversation with your provider.
Written by Michelle Mullins, AGACNP-BC, FNP-BC
Last reviewed/updated: August 14, 2026