The FDA announced today that it will remove the boxed warnings on menopausal hormone therapies for cardiovascular disease, breast cancer and probable dementia. The warning about endometrial cancer stays on estrogen-alone products, where it belongs. The decision follows the expert panel we wrote about in July, a review of the evidence and a comment period. Alongside it, the agency approved a generic version of the most common conjugated-estrogen product and a non-hormonal treatment for hot flashes. For two decades, every woman considering hormone therapy read a label built on the 2002 Women's Health Initiative, whose participants averaged 63 years old and took a formulation now rarely used. The label is finally catching up to the evidence about the women who actually ask for the drug.
What the evidence supports, and what it doesn't. Reanalyses of the original trial and later studies support a timing hypothesis: women who start therapy before 60 or within about ten years of menopause see a different balance of benefits and risks — relief of symptoms, protection against bone loss, and in some analyses no excess cardiovascular risk — than women who start later, for whom the 2002 risks remain real. Removing a boxed warning does not make a drug safe for everyone; it removes a blanket statement that was wrong for most of the people reading it. Risks of clots, stroke, and hormone-sensitive cancers vary by age, formulation, route and history, which is exactly what a label cannot capture and a conversation can.
Loree’s takeI have had this conversation with hundreds of women, and for twenty years it began with me explaining why the scariest box on the label did not describe them. Today it gets shorter. It does not get simpler: the right decision still depends on her age, her symptoms, her family history and what she is willing to trade. The warning was a substitute for that conversation. Its removal is an invitation to have it.
Ben’s takeTwo things happened this year that look opposite and are the same. The FDA took a warning off a drug when the evidence said the warning was built on the wrong population. The longevity industry, all year, kept putting claims on products when the evidence was built on the wrong species. The regulator corrected in twenty-two years. The industry does not correct, because the mouse paper is the product. Hold both to the same question: whose data is this, and am I them?
Why it matters to a recovering optimizer. This is a story about one label, but it is the clearest example we have of the sex-differences problem in medicine: a trial of one group of women was applied to all of them for a generation. The same problem runs through most longevity research, which is still done mostly in men and in animals. Whenever a longevity claim reaches you, ask what population generated it. Today the FDA asked that question about its own label and changed the answer.
A warning is not a verdict about you. Neither is a promise. Ask whose data it came from.
This post reports on evidence and expert disagreement. It is not medical advice; decisions about treatment belong with you and your clinician.