HRT Safety: What the WHI Study Actually Found, 20 Years On

The 2002 headlines reshaped menopause care for a generation. Here is what the data showed, what was misread, and where the guidance stands now.

In July 2002 the Women's Health Initiative halted one arm of its hormone therapy trial, and the coverage that followed cut hormone therapy prescribing by more than half within a year. Two decades of reanalysis have substantially revised that interpretation — but the original headlines still shape how many women, and many clinicians, think about HRT.

What the trial actually studied

The WHI enrolled women with an average age of 63, more than a decade past menopause on average, and used two specific formulations: conjugated equine estrogens and medroxyprogesterone acetate, a synthetic progestin. It was designed to test whether hormone therapy prevented chronic disease in older women, not whether it safely relieved symptoms in women in their early fifties. Most women seeking HRT today do not resemble the trial population, and most are not offered the trial's formulations.

What it found

The combined estrogen-progestin arm showed a small absolute increase in breast cancer, stroke, and venous thromboembolism, alongside reductions in fracture and colorectal cancer. The estrogen-only arm — women who had had a hysterectomy — showed no increase in breast cancer, and long-term follow-up has suggested a possible reduction. That distinction was largely absent from the original coverage.

The timing hypothesis

Subsequent age-stratified analyses found that women who started hormone therapy under 60, or within ten years of their final period, had a materially different risk profile from those who started later — including a neutral-to-favourable effect on cardiovascular outcomes. This "timing hypothesis" now underpins the guidance from the Menopause Society and equivalent bodies internationally.

Where guidance stands now

For healthy symptomatic women under 60 and within ten years of menopause, the benefits of hormone therapy generally outweigh the risks. Transdermal estrogen avoids the clot signal associated with oral routes. Micronized progesterone appears to carry a more favourable profile than the synthetic progestin used in the trial. The arbitrary five-year stopping rule that followed 2002 is no longer recommended; duration is a decision to revisit annually rather than a deadline.

What has not changed

Hormone therapy is not appropriate for everyone. Active or recent breast cancer, unexplained vaginal bleeding, active liver disease, and a history of estrogen-dependent cancer or unprovoked clotting all require individual clinical evaluation. Risk is personal, and absolute risk — not relative risk — is the number worth discussing with your clinician. Our menopause HRT guide covers how the current evidence is applied in practice.