Menopausal hormone therapy tracked with less Alzheimer pathology at autopsy

Design

retrospective cohort study across two independent datasets, with neuropathological and biomarker outcomes

Population

258 hormone therapy users and 2,701 non-users with autopsy data; 110 and 1,948 in the imaging cohort

Primary outcome

raised Alzheimer disease pathology at autopsy

Effect

odds ratio 0.65 (95% CI 0.48–0.88, p=0.005); clinical dementia diagnosis OR 0.61 (0.55–0.67)

Whether the fall in oestrogen at menopause contributes to women's higher Alzheimer risk — and whether replacing it changes anything — has been argued for thirty years without resolution. This cohort study looked at two independent datasets: the National Alzheimer's Coordinating Center, with neuropathological data on 258 women who had used estrogen-only hormone therapy and 2,701 who had not, and the Alzheimer's Disease Neuroimaging Initiative, with 110 users and 1,948 non-users.

The primary outcome was Alzheimer pathology at autopsy, and the odds of raised pathology were lower in hormone therapy users (odds ratio 0.65, 95% CI 0.48 to 0.88, p=0.005). Secondary outcomes moved the same way: lower amyloid burden measured in plasma and in cerebrospinal fluid, lower odds of a clinical dementia diagnosis (0.61, 0.55 to 0.67) and lower odds of reported memory or functional decline (0.67, 0.61 to 0.74).

The authors are explicit that this does not address causality, and the reasons matter in clinic. Hormone therapy use was self-reported, the design retrospective, and women who took it in later life differ systematically from those who did not — in education, healthcare access and baseline health. The finding is worth knowing when a woman asks whether her past hormone therapy helped or harmed her brain; it is not a reason to prescribe it for that purpose, and no trial has tested it as a dementia intervention with these endpoints.

• Do not prescribe menopausal hormone therapy for dementia prevention on this evidence
• Record past hormone therapy use in a cognitive history — it is rarely asked about
• Note this studied estrogen-only therapy, not combined preparations
• Distinguish for the patient between an autopsy association and a treatment effect
• Continue to weigh hormone therapy on its established indications and risks

Why it matters

It reopens a question most clinicians treat as settled against hormone therapy, using pathology rather than symptoms as the endpoint.

Don't overread it

This was observational with self-reported exposure — it cannot show that taking hormone therapy prevents Alzheimer pathology.

The statistics, in plain English

An odds ratio of 0.65 for autopsy pathology, with an interval of 0.48 to 0.88, is a modest association that stays clear of 1.0 — but the confidence interval measures sampling error, not bias, and bias is the problem here. Women who used hormone therapy and later came to autopsy in a research cohort are a selected group twice over. The clinical dementia estimate of 0.61 has a much tighter interval because it draws on far more people, which makes it more precise without making it less confounded.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.