For decades, the relationship between estrogen and Alzheimer's disease has been one of the most contested questions in women's health — complicated by the Women's Health Initiative fallout, timing debates, and conflicting observational data. This large dual-cohort study shifts the conversation meaningfully by anchoring findings not just in clinical diagnosis, but in actual neuropathological confirmation at autopsy and validated biomarkers.
Drawing on two independent datasets — the National Alzheimer's Coordinating Center (NACC) and the Alzheimer's Disease Neuroimaging Initiative (ADNI) — the analysis examined estrogen-only menopausal hormone therapy (MHT) in nearly 5,000 women aged 50 and older. Among the 258 NACC participants with autopsy data, MHT users showed a 35% reduction in odds of significant Alzheimer's neuropathology compared with non-users (OR 0.65, 95% CI 0.48–0.88, p = 0.005). Secondary outcomes reinforced this signal: MHT use was independently associated with reduced amyloid burden on imaging and fluid biomarkers, offering mechanistic plausibility beyond the primary autopsy result.
What distinguishes this work from prior observational studies is the neuropathological endpoint. Clinical diagnoses of dementia carry well-documented misclassification risk; autopsy-confirmed amyloid and tau pathology is far closer to biological ground truth. This elevates the findings above typical registry-based analyses. That said, critical limitations remain. This is still observational data — women who chose estrogen-only therapy likely differ systematically from non-users in health behaviors, socioeconomic status, and baseline cardiovascular risk. The estrogen-only formulation is relevant here, as combined estrogen-progestin regimens carry distinct risk profiles that may not replicate this benefit. The 'critical window' or timing hypothesis — suggesting early post-menopausal initiation matters — cannot be fully resolved with this cohort design. Whether this represents true neuroprotection or healthy-user selection bias demands a prospective randomized trial. Incremental but directionally important, this study adds the strongest pathological evidence yet that estrogen decline and AD risk in women are biologically linked.