Autopsies, living brain scans, and 183,000 UK Biobank women converge on one finding: menopausal hormone therapy may reduce Alzheimer's risk when started between ages 46 and 56
Autopsies, living brain scans, and 183,000 UK Biobank women converge on one finding: menopausal hormone therapy may reduce Alzheimer's risk when started between ages 46 and 56
A Stanford study using postmortem brain data (Neurology, August 2026) found 35% lower odds of Alzheimer's pathology in women on estrogen therapy compared with those who never used it. A study of 183,450 UK Biobank women (Alzheimer's & Dementia, August 2026) linked the therapy to lower dementia risk, with the strongest benefit in women who started treatment between ages 46 and 56. Both findings aligned with an earlier result: the therapy may neutralize the risk conferred by APOEε4, the strongest known genetic risk factor for Alzheimer's disease.
Twenty years ago, the Women's Health Initiative Memory Study (WHIMS) found that women aged 65 and older taking estrogen plus progestin faced nearly double the risk of probable dementia. The share of U.S. women using such therapy dropped from 27% to less than 5%. Yet in the estrogen-only arm of WHIMS no such harm appeared; the effect was weaker and statistically nonsignificant. This detail was overshadowed when the finding was generalized to all hormone therapy.
On August 12, 2026, a Stanford team tested this association for the first time in postmortem brain tissue. Among 21,462 participants, autopsy data were available for 258 women on estrogen alone and 2,701 women who received no therapy. The odds of Alzheimer's pathology at autopsy were 35% lower in the estrogen group. "We examined every standard for diagnosing Alzheimer's, including the gold standard, signs of the disease in autopsied brains," says senior author Hadi Hosseini. Co-author Jennifer Bruno calls the effect "modest but significant": smaller than the effect of age or APOEε4, but consistent across measures.
The second study, published the same month in Alzheimer's & Dementia, followed 183,450 UK Biobank women over 13 years of observation. Among those who used the therapy, dementia risk was 10% lower and Alzheimer's risk 16% lower. The benefit was larger in women who had undergone oophorectomy (26% lower dementia risk) and in APOEε4 carriers (13% lower risk). The age at which treatment began proved decisive: protection appeared only when therapy started between ages 46 and 56, shortly after menopause, not earlier and not later.
The "critical window hypothesis," proposed in 2013, predicted this pattern: the brain remains responsive to estrogen in the first years after menopause, but roughly ten years later the window closes. This timing also explains the discrepancy with WHIMS, where therapy began at age 65 or older, one and a half to two decades past the protective window. The same WHIMS puzzle has a second explanation: two days earlier, an analysis of hormone therapy formulations linked the protection seen in APOEε4 carriers to a specific estrogen, estradiol.
A 2024 study in living women, using positron emission tomography (PET, which measures brain metabolism) and blood analysis, may have identified the mechanism. Among those who had not used hormone therapy, carriers of APOEε4 (which raises Alzheimer's risk more in women than in men) showed higher levels of tau protein, a marker of memory decline, compared with non-carriers. Among those who had used the therapy, this difference was absent, as though estrogen erased the risk.
Autopsy, living brain imaging, and population statistics detect signs of Alzheimer's in different ways and are prone to different errors; their convergence in direction, and specifically on the APOEε4 detail, carries more weight than any single result alone. In 2025, the FDA removed the black-box warning introduced after WHIMS from some of these medications, a regulatory move in the same direction as the science.