Menopausal hormone therapy using estrogen alone appears to be associated with a meaningfully lower risk of Alzheimer’s disease, according to new research from Stanford Medicine that examined the strongest form of evidence available: what pathologists actually find in the brain after death.

The study, published August 12 in the journal Neurology, arrives at a conclusion that runs against a widely held assumption. For more than two decades, much of the medical field has advised against using hormone therapy with cognitive protection in mind.

Why This Study Is Different

Previous research on this question has relied largely on clinical diagnoses or on biomarkers drawn from blood and cerebrospinal fluid.

Senior author Hadi Hosseini, an associate professor of psychiatry and behavioral sciences, said the distinguishing feature of this work is that it examined all the standard measures of Alzheimer’s diagnosis including the definitive one, the pathological hallmarks visible in autopsied brains. He noted that no prior study had assessed postmortem outcomes in substantial numbers.

His reasoning is practical. Many conditions affect memory, and clinical diagnoses are frequently inaccurate. Biomarkers have improved considerably, but they cannot indicate how much damage exists or precisely where it sits. Examining the brain itself shows both.

Jennifer Bruno, an instructor in the same department, and former postdoctoral scholar Jacob Shaw share lead authorship.

What the Researchers Did

The team drew on two existing databases containing 21,462 participants in total.

From that pool they analyzed 258 brain autopsies from women who reported using estrogen-only hormone therapy, comparing them against roughly 2,701 women who reported no use of any hormone therapy.

The comparison focused on three features: amyloid plaques, neurofibrillary tangles, and the density of individual plaques. Levels and locations were scored and compared across groups.

The Findings

Estrogen-only therapy use was associated with a statistically significant reduction in these Alzheimer’s hallmarks.

Expressed as an odds ratio, the analysis found a 35 percent reduction in the odds of developing Alzheimer’s pathology among users.

Data from living participants supported the autopsy results. Estrogen-only users had 39 percent lower odds of receiving a clinical dementia diagnosis during their lifetimes, performed better on memory testing, and maintained greater ability to function independently.

Bruno characterized the effect as modest but meaningful. She was careful to place it in context, noting it is nowhere near as powerful as established factors such as age, carrying the APOE4 genetic variant, or a history of hypertension. What impressed the researchers was consistency across multiple different Alzheimer’s-related outcomes.

The analysis adjusted for those risk factors along with education and race.

Who This Actually Applies To

Estrogen-only therapy is generally prescribed to women who no longer have a uterus, since estrogen without progestin raises endometrial cancer risk in women who do.

That makes the population fairly specific but far from small. By age 60, more than 30 percent of women have had a hysterectomy. Estrogen-only treatment is the recommended formulation for them.

Participants in the study skewed older, averaging around 70 years of age, and many had undergone hysterectomies, which is why most were on estrogen-only regimens.

Important Limitations

The researchers were direct about what the study cannot tell us.

The number of autopsied women using estrogen-plus-progestin formulations was too small to draw any conclusion about that combination.

Topical estrogen users were not included, only those on oral therapy.

The findings therefore apply specifically to oral estrogen-only regimens. They do not extend to combination therapy or to topical administration.

Association is also not causation. The link is clear in the data, but the study design cannot establish that the therapy caused the reduction.

Why Earlier Research Conflicted

Hosseini offered a thorough explanation for why this field has produced such contradictory results.

Studies have differed in the age at which participants began therapy, in which clinical outcomes they measured, in which biomarkers they used, and in whether they grouped different formulations together. Routes of administration and duration of use have varied. Some studies did not adequately adjust for genetic, cardiovascular, or other risk factors.

Bruno added that assessments of cognitive decline carry inherent subjectivity. Results can depend on who performs the evaluation, and apparent decline may stem from causes ranging from a stroke to simply a poor night’s sleep.

The 2003 Turning Point

The historical context matters enormously here.

Early work suggested hormone therapy might protect against dementia. Subsequent studies proved inconclusive.

Then in 2003 the Women’s Health Initiative Memory Study reported the opposite finding, indicating that estrogen-plus-progestin formulations appeared to increase dementia risk, at least when started at older ages. The same initiative linked that combination to elevated breast cancer and heart disease risk.

The FDA removed a black-box warning to that effect in 2025. By then the damage to public confidence was long established.

Usage figures tell the story. Lifetime use among women was once estimated near 27 percent. Hosseini said that figure has since fallen below 5 percent.

Timing May Matter

One additional point from the researchers deserves attention.

Recent literature suggests hormone therapy is more beneficial when started during or shortly after menopause rather than years later.

The women in this study tended to be older, meaning many started late. Hosseini suggested that if anything, this would cause the study to underestimate the benefit rather than overstate it.

A Direct Challenge to Guidance

Bruno framed the significance plainly. For a long stretch, the standard recommendation was not to use hormone therapy for memory decline. This study, in her words, runs directly counter to that advice.

Whether guidelines shift will depend on replication and on whether prospective trials support the association.

What This Means Practically

For anyone weighing this personally, a few things are worth holding onto.

The findings apply narrowly to oral estrogen-only therapy in women without a uterus.

The effect size is modest compared with major risk factors.

Hormone therapy carries its own risk profile that must be weighed individually against benefits.

This is one study, however methodologically strong, and it contradicts a body of earlier work.

Decisions about hormone therapy involve multiple considerations beyond cognition, including bone health, cardiovascular risk, cancer risk, and symptom relief. Those trade-offs vary considerably from person to person, which makes this a conversation for a doctor who knows your history rather than something to act on from a headline.

The research was funded by the National Institutes of Health.