Guide · Reviewed August 14, 2026
HRT and Alzheimer's risk: what the 2026 study found, and what it did not
A cohort study published in Neurology found that women who reported using estrogen-only menopausal hormone therapy had lower odds of Alzheimer pathology at autopsy, an odds ratio of 0.65 (95% CI 0.48-0.88, p = 0.005). It was published online August 12, 2026, in the September 8, 2026 issue. The same analysis reported lower odds of a clinical dementia diagnosis (OR 0.61, 95% CI 0.55-0.67) and of memory or functional decline (OR 0.67, 95% CI 0.61-0.74).
The authors call these small but significant associations, and they state that the results do not address causality. The design is retrospective, which limits how far the findings carry. The exposure studied was estrogen-only therapy, not the combined estrogen-plus-progestogen regimen most women with a uterus take. The FDA labeling continues to state that menopausal hormone therapy is not approved to prevent dementia.
Medical disclaimer: this page is independent research and journalism, not medical advice. Decisions about hormone therapy belong with a licensed clinician who knows your history.
What the headline will say, and what the study tested
A result like this travels as "HRT protects against Alzheimer's." The paper does not say that, and the gap between the two is where a reader gets misled. Here is each claim you are likely to encounter set against what the cohort actually measured.
| The headline version | What the paper supports |
|---|---|
| HRT prevents Alzheimer's. | An association, not prevention. The authors state the results do not address causality. |
| HRT cuts Alzheimer's risk by 35%. | An odds ratio of 0.65 for increased Alzheimer pathology on autopsy, with a confidence interval running from 0.48 to 0.88. The authors describe the associations as small. |
| HRT protects the brain. | Estrogen-only therapy was the exposure. Combined estrogen-plus-progestogen therapy was not tested here. |
| A reason to start HRT. | Not on this evidence. FDA labeling states menopausal hormone therapy is not approved to prevent dementia. |
What did the 2026 Neurology study actually measure?
Two independent data sets, with different outcomes in each. Neuropathology came from the National Alzheimer's Coordinating Center, comparing 258 women who reported estrogen-only HRT use (mean age at death 81.9) against 2,701 who reported no use (mean age at death 82.2). Biomarker and clinical outcomes came from the Alzheimer's Disease Neuroimaging Initiative, comparing 110 users (mean age 76.5) against 1,948 non-users (mean age 73.2).
The primary outcome was the odds of increased Alzheimer pathology on autopsy. The reported figures:
- Increased Alzheimer pathology at autopsy: OR 0.65 (95% CI 0.48-0.88, p = 0.005)
- Amyloid load measured in plasma: β = 0.44 (95% CI 0.16-0.73, p = 0.0025)
- Amyloid load measured in cerebrospinal fluid: β = 0.07 (95% CI 0.002-0.13, p = 0.030)
- Clinical dementia diagnosis: OR 0.61(95% CI 0.55-0.67, p < 0.0001)
- Symptoms of memory or functional decline: OR 0.67(95% CI 0.61-0.74, p < 0.0001)
Source: Bruno JJ, Shaw JS, Hosseini SMH, for the Alzheimer's Disease Neuroimaging Initiative, Association Between Menopausal Hormone Therapy and Alzheimer Disease Neuropathology, Neurology 2026;107(5):e218413, doi 10.1212/WNL.0000000000218413, epub August 12, 2026.
Why an autopsy cohort cannot answer the causal question
Participants reported their own HRT use, and the researchers looked at what happened afterwards. Nobody was assigned to take estrogen or not. Women who took HRT in that era differed from women who did not in ways that also track with dementia risk, including access to care, education, and general health at the point of prescription. That is the standard limitation of observational work, and it is why the authors write that their results do not address causality and cite limited generalizability from the retrospective design.
The accompanying editorial in the same issue frames the open questions as "When, Which, and for Whom?" (Salemme S, Neurology 2026;107(5):e218503, doi 10.1212/WNL.0000000000218503). Timing, formulation, and patient selection are unresolved, which is a fair reading of where the evidence sits.
How this sits against the FDA label
In February 2026 the FDA approved labeling changes removing the boxed warning from six menopausal hormone therapy products, deleting the cardiovascular-disease, breast-cancer, and probable-dementia statements. (FDA) Two separate things follow from that, and they are easy to run together.
Removing the probable-dementia statement corrected an overstatement of harm drawn largely from a 2002 trial of much older women. It did not add a benefit claim. The labeling still states these products are not approved to prevent dementia, so a clinician has no label basis for prescribing HRT as a cognitive prophylactic. The full account of what came off the label and what remains is in our black box warning guide.
So should you start HRT for brain protection?
Not on this evidence. The established reasons to take HRT are menopausal symptoms and bone protection, and the timing rule that governs the risk-benefit balance is about starting within roughly ten years of menopause rather than about dementia. If you are weighing HRT for symptoms and this finding is one input among several, that is a reasonable way to hold it. If it is the reason, the study does not carry that weight.
What to do with it in a consultation: ask which formulation is being proposed, since this cohort tested estrogen-only therapy, and ask how your own timing and history change the balance. Our 2026 safety guide covers the risk picture in full, and the perimenopause hub covers starting decisions for the younger cohort.
Common questions
Does HRT prevent Alzheimer's disease?
No study has shown that, and this one does not either. The August 2026 Neurology cohort found that women who reported using estrogen-only menopausal hormone therapy had lower odds of Alzheimer pathology at autopsy (odds ratio 0.65, 95% CI 0.48-0.88, p = 0.005). The authors describe these as small but significant associations and state plainly that their results do not address causality. An association found by looking backwards at who happened to take estrogen is not evidence that taking it causes the outcome.
What did the 2026 Neurology study actually measure?
Two things, in two separate cohorts. Neuropathology at autopsy came from the National Alzheimer's Coordinating Center: 258 estrogen-only HRT users (mean age at death 81.9) against 2,701 non-users (mean age at death 82.2). Biomarkers and clinical outcomes came from the Alzheimer's Disease Neuroimaging Initiative: 110 users (mean age 76.5) against 1,948 non-users (mean age 73.2). The primary outcome was the odds of increased Alzheimer pathology on autopsy.
Was combined estrogen-plus-progestogen HRT included?
No. The exposure tested was estrogen-only menopausal hormone therapy. Most women with a uterus are prescribed a progestogen alongside systemic estrogen to protect the uterine lining, so the regimen most commonly prescribed in that group was not what this study measured. Applying the finding to combined therapy goes beyond the data.
Should I start HRT to protect my brain?
That is not a decision this study supports, and the FDA labeling is explicit that menopausal hormone therapy is not approved to prevent dementia. The February 2026 label changes removed the probable-dementia risk statement from the boxed warning, which corrected an overstatement of harm. Removing a warning is not an approval for prevention. Discuss HRT with a clinician on its established indications, which are menopausal symptoms and bone protection.
Why does the removal of the dementia warning not settle this?
The boxed warning carried a probable-dementia risk statement drawn largely from a 2002 trial of much older women, and the FDA removed that statement from six menopausal hormone therapy products in February 2026. That answers whether the old warning overstated harm. It does not answer whether HRT protects the brain, which is a separate question that the 2026 Neurology cohort addresses with association data rather than a trial.
Sources
- Bruno JJ, Shaw JS, Hosseini SMH, for the Alzheimer's Disease Neuroimaging Initiative. Association Between Menopausal Hormone Therapy and Alzheimer Disease Neuropathology. Neurology 2026;107(5):e218413. doi 10.1212/WNL.0000000000218413. PubMed 42585606
- Salemme S. Menopausal Hormone Therapy and Alzheimer Disease: When, Which, and for Whom? Neurology 2026;107(5):e218503. doi 10.1212/WNL.0000000000218503. PubMed 42585609
- FDA. FDA Approves Labeling Changes for Menopausal Hormone Therapy Products. fda.gov
Written and reviewed by Iacob Pastina. Every figure on this page was read from the published records cited above on August 14, 2026.
Get a monthly read on the HRT market.
Pricing changes, formulary updates, shortage status. One short email per month. Unsubscribe anytime.
No spam. We never sell your email. Editorial policy.
Midi Health
Sesame Care