What the WHI study actually showed about HRT and breast cancer

The headline that changed everything

In July 2002, one trial reversed three decades of practice. The Women’s Health Initiative (WHI) — a large, well-run, publicly funded randomised trial — was stopped early and reported that combined hormone therapy (oestrogen plus progestin) caused more breast cancer and more heart disease. The published conclusion was blunt:

“Overall health risks exceeded benefits … for an average 5.2-year follow-up among healthy postmenopausal US women.”

Prescriptions collapsed worldwide almost overnight. A generation of women either stopped HRT or were never offered it. The problem is not what the trial found — it is what the headlines did with it.

This page is a forensic appraisal: a structured comparison of what the WHI data actually show against what the abstract claimed and what the public took away. It is the same method applied to any single trial — read the numbers, separate relative from absolute, and check whether the conclusion is calibrated to the people who were actually studied.

What the trial actually found, in absolute terms

To its credit, the WHI reported its results not only as hazard ratios but also as absolute risks per 10,000 women per year. Those absolute numbers are small — and they are the numbers the public conversation lost.

Outcome	Hazard ratio (nominal CI)	Adjusted CI	Absolute effect per 10,000 women/year
Coronary heart disease	1.29 (1.02–1.63)	0.85–1.97 (not significant)	+7 events
Stroke	1.41 (1.07–1.85)	0.86–2.31 (not significant)	+8 events
Pulmonary embolism	2.13 (1.39–3.25)	0.99–4.56 (not significant)	+8 events
Invasive breast cancer	1.26 (1.00–1.59)	0.83–1.92 (not significant)	+8 cancers
Colorectal cancer	0.63 (0.43–0.92)	0.32–1.24 (not significant)	−6 cancers
Hip fracture	0.66 (0.45–0.98)	0.33–1.33 (not significant)	−5 fractures

Hazard ratios, confidence intervals and absolute risks from Rossouw et al., JAMA 2002 and the JAMA full text.

The breast-cancer signal that triggered a global prescribing collapse was 8 extra cases per 10,000 women per year — an absolute increase of 0.08% per year, fewer than one extra case per thousand women each year. The famous “26% increase” is the relative way of describing those same 8 cases. Both are true; only one of them tells you how much your own risk changes.

The statistical detail the headlines dropped

The WHI tested seven main outcomes at once. When you test many outcomes, some will look significant by chance, so the trialists correctly reported adjusted confidence intervals alongside the headline nominal ones. On those adjusted intervals, breast cancer, coronary heart disease, stroke and pulmonary embolism were each no longer statistically significant — only the combined “global index” survived (JAMA full text).

In other words, the “26% more breast cancer / 29% more heart disease” figures rest on the nominal numbers that the trialists themselves flagged as uncorrected. That nuance did not make it into the press release.

Who was actually in the study

The average WHI participant was 63 years old, and about 21% were aged 70–79 (baseline data). Only around 4% had moderate-to-severe hot flushes or night sweats at enrolment (Hays et al., NEJM 2003). The average woman was roughly 13 years past menopause.

This was a trial of chronic-disease prevention in older women — not a trial of symptomatic women in their early 50s. The fear it generated was then applied to exactly the women it never studied: those near menopause seeking relief from hot flushes, sleep disruption and mood change.

Was the conclusion supported?

On a structured read of trial design, conduct, analysis and reporting, the WHI comes out largely sound — which is itself the finding:

• 🟢 Honest reporting. Hard clinical endpoints, intention-to-treat analysis, harms reported with the same rigour as benefits, absolute risks published alongside relative ones, both nominal and adjusted statistics shown. Little of the spin seen in industry-run trials.
• 🟠 An over-general conclusion. “Healthy postmenopausal women” carried no age or timing caveat, which invited the leap to all menopausal women. The 2007 reanalysis later had to supply the missing nuance.
• 🟠 A composite that drove the headline. The only result significant on adjusted statistics was a heterogeneous “global index” that bundled a small cancer signal with a small fracture benefit as if they were equivalent.

The distortion was largely a press and guideline phenomenon, not a flaw baked into the paper. The trial answered the question it asked; the world answered a different question with its results.

What Australian guidance says now

• The Australasian Menopause Society advises starting HRT within 10 years of menopause, does not recommend it for preventing heart disease, and notes no evidence of cardiovascular harm for women starting close to menopause — with a trend toward benefit.
• The 2007 WHI reanalysis found coronary heart disease hazard ratios of 0.76 for women less than 10 years past menopause versus 1.28 for those 20 or more years out — the evidence behind the “window of opportunity”.
• RACGP and Australian Prescriber align: for healthy women under 60 or within 10 years of menopause, the benefits of HRT generally outweigh the risks, and the WHI drug was never approved for cardiovascular prevention in the first place.
• The Cochrane review agrees on direction but rates the certainty lower than the 2002 headlines implied: combined therapy “probably makes little to no difference to the risk of a coronary event” and “probably increases breast cancer”.

What this means if you are weighing up HRT

The WHI does not mean HRT is unsafe, and it does not mean it is risk-free. It means the real risks are small, absolute, and dependent on your age, your timing and the type of therapy — and they sit on a scale alongside the benefits you are actually seeking. The relative-risk headline that frightened a generation was never the number that tells an individual woman how much her own risk changes.

The decision is one to make with your own GP, using your own history. The point of reading the evidence this way is to walk into that conversation with the absolute numbers, not the headline.

What this article is and is not

This is general health information and an analysis of the published medical literature, drawn from Australian primary-tier sources — the Australasian Menopause Society, RACGP and Australian Prescriber — and the original trial publications. It is not personal medical advice and does not create a doctor–patient relationship. Decisions about hormone therapy depend on your individual circumstances, which only your own doctor knows.

For a broader overview of menopause management, see the menopause and perimenopause framework. For Australian consumer resources: Australasian Menopause Society, Jean Hailes for Women’s Health, HealthDirect — Menopause.

---

Sources cited

1. Rossouw et al. Risks and benefits of combined oestrogen–progestin therapy in healthy postmenopausal women (WHI). JAMA 2002 (PMID 12117397)
2. WHI 2002 — JAMA full text
3. Rossouw et al. Postmenopausal hormone therapy by age and years since menopause. JAMA 2007
4. Hays et al. Quality of life and combined HRT. NEJM 2003
5. Cochrane CD004143 (Marjoribanks 2017)
6. Australasian Menopause Society — Risks and benefits of MHT/HRT
7. Australian Prescriber — Conjugated oestrogens and medroxyprogesterone acetate
8. RACGP — Making choices at menopause
9. ClinicalTrials.gov — NCT00000611