What the New Menopause Brain Studies Do (and Don't) Show

Short answer. One 2026 study found less grey matter (the part of the brain made of nerve cells) in some brain regions after menopause (Zühlsdorff et al., 2026). Another study found blood proteins that are higher after menopause and are linked to brain aging and Alzheimer's dementia risk (Wood Alexander et al., 2026). Neither study gave anyone a treatment, so neither shows that menopause causes dementia (Zühlsdorff et al., 2026; Wood Alexander et al., 2026). A senior author of the UCSF study said in the university's news release that the team doesn't think menopause directly causes dementia.

Key facts

• Cambridge researchers looked at about 125,000 UK Biobank women, about 11,000 of them with brain scans. Women after menopause had less grey matter than women before menopause in the hippocampus, entorhinal cortex, and anterior cingulate cortex (Zühlsdorff et al., 2026).
• A UCSF and University of Toronto study found 16 blood proteins that are higher after menopause. In older women, a higher menopause protein score went along with worse thinking and memory outcomes and, in the UK Biobank, a 15% higher risk of Alzheimer's dementia (Wood Alexander et al., 2026).
• Hormone therapy didn't seem to reverse the differences in grey matter, mood or sleep. Women who used it also had more mental health symptoms before they started treatment (Zühlsdorff et al., 2026).

One woman wrote in an online forum: “I don't know how much of this is perimenopause or my regular clinical depression.”

If you've wondered what's behind how you feel in midlife, these two new studies offer a first sketch of what might be going on, though many questions are still open.

The Cambridge brain-scan study

Published in Psychological Medicine, the study compared three groups of women: women before menopause, women after menopause who never used hormone replacement therapy (HRT), and women after menopause who did (Zühlsdorff et al., 2026). Both groups of postmenopausal women had less grey matter in the three regions above. The hippocampus and anterior cingulate cortex were smallest in women who used HRT.

Among the women with brain scans, only 762 of 10,873 were still premenopausal. Postmenopausal women also said they had more anxiety, depression, and sleep difficulties.

On the prospective memory and digit-span tasks, no differences between the groups held up once the researchers corrected for having run many tests. Reaction time was an exception, though the paper is inconsistent about which postmenopausal group differed.

Women who used HRT said they had more anxiety and depression than postmenopausal women who didn't. Further analyses, run after the main results, found they already had more symptoms before treatment, which may be why a doctor prescribed it. In the university's press release, one author said the brain regions with differences tend to be affected by Alzheimer's disease.

The UCSF and Toronto blood-protein study

The Nature Medicine study measured more than 100 brain-related proteins in the blood of 80 women aged 43 to 58 and 36 men. Of the women, 30 were premenopausal, 26 perimenopausal and 24 postmenopausal. Sixteen proteins were higher in postmenopausal women, and the changes followed hormone levels more closely than they followed age. When the team repeated the check in 2,814 UK Biobank women of similar ages, they found similar changes (Wood Alexander et al., 2026).

In four groups of older women (11,925 in total), higher menopause protein scores went along with worse thinking and memory outcomes. In the UK Biobank, higher scores were linked to a 15% higher risk of Alzheimer's dementia over about 16 years of follow-up. The link to dementia of any cause wasn't strong enough to rule out chance (Wood Alexander et al., 2026; UCSF news release, 2026).

How do the two studies compare?

A study that followed the same women over time

The Cambridge study compared different women at one point in time. A 2026 study in Nature Communications followed the same women over time instead. It compared puberty, pregnancy and menopause in 1,095 women, with control groups whose menopause status stayed the same. In this study, women who went through the menopausal transition showed no loss of brain volume, while the control groups who stayed premenopausal or postmenopausal did. The authors describe menopause as marked by the absence of volume loss rather than faster loss (van ’t Hof et al., 2026). The two designs answer different questions, so this doesn't cancel the Cambridge result, but it's a reason to be careful about reading a one-time comparison as the effect of menopause itself.

Does this mean menopause causes dementia?

These studies don't show that. A senior UCSF author said in the university's news release that the team doesn't think menopause directly causes dementia. The proteins may help predict risk years later (Wood Alexander et al., 2026). The starting group of women was small, and the proteins were measured at one point in time, so they can't show what came first. Symptoms came from women's own reports.

The 2026 FDA decision on hormone therapy

On February 12, 2026, the US Food and Drug Administration (FDA) approved new labels for the first six menopausal hormone therapy products. The new labels drop three risk statements from the boxed warning: cardiovascular disease, breast cancer, and probable dementia. In all, 29 companies have sent in proposed label changes (FDA, 2026).

The decision is about what labels say. It doesn't test whether hormone therapy protects or changes the brain. The Cambridge study above found no sign that it reversed the grey matter differences (Zühlsdorff et al., 2026).

One randomized trial (chance decides who gets the treatment) followed 172 women aged 45 to 60 who started with no depressive symptoms. Over 12 months, 32.3% of women on placebo scored 16 or higher on a depression questionnaire (CES-D) at least once, compared with 17.3% of women using an estradiol patch (estradiol is a form of estrogen) plus periodic progesterone (Gordon et al., 2018). The mood benefit was clearest early in the menopause transition. This is a mood result from a questionnaire, not a diagnosis and not a brain-imaging result. A trial and an observational study like the Cambridge one can look different, because the trial assigned the treatment by chance and the observational study couldn't.

What the research can’t tell us

• Both studies are observational: they looked at women as they already were, and nobody was assigned a treatment.
• The Cambridge analysis compares different women at one point in time, so it can't show that menopause caused the differences. The analyses took age into account, but a letter to the journal (Keye, 2026) says other differences between women may remain. The authors replied that they report links, not causes. UK Biobank volunteers are also healthier and less ethnically diverse than the UK population.
• The UCSF group of midlife women is small (80 women).
• The research doesn't yet say what to do with these markers. The team plans a blood test and a long-term study, the Longitudinal Menopause Project.

When should you talk to a professional?

See a clinician who knows menopause if memory or concentration problems get in the way of your life, or if you're weighing hormone therapy. Brain fog has many causes, and menopause is only one of them.

Where Soula can fit in

Notes on what you notice can help before a doctor's visit. Soula has cycle tracking, check-ins and an AI-assistant chat, which can help you put things into words so you aren't relying on memory in the appointment. It isn't a diagnostic tool, and it encourages you to take what you notice to a clinician.

FAQ

What did the Cambridge study find about thinking and memory?

In one large study, postmenopausal women said they had more anxiety, depression, and sleep trouble, and they had less grey matter in three brain regions. On two memory tasks, no differences between groups held up once results were corrected for running many tests. Brain fog itself wasn't measured (Zühlsdorff et al., 2026).

Did the FDA's 2026 change mean hormone therapy is safe for the brain?

No. The change updated warning labels and didn't test brain outcomes. Whether hormone therapy is right for you is something to decide with a clinician (FDA, 2026).

Does menopause shrink the brain?

Women after menopause had less grey matter in three brain regions in one UK Biobank analysis. That study compared different women at one point in time, though, and a 2026 study that followed the same women over time reported no brain volume loss in women going through the transition (Zühlsdorff et al., 2026; van ’t Hof et al., 2026).

Does menopause cause dementia?

Not on this evidence. Higher protein scores went along with Alzheimer's dementia risk in the UK Biobank, and the proteins may help predict risk years later, but a senior UCSF author said the team doesn't think menopause is directly causing dementia (UCSF news release, 2026; Wood Alexander et al., 2026).

Does hormone therapy protect the brain?

In the Cambridge study, hormone therapy didn't seem to reverse the differences in grey matter, mood, or sleep. Women who used it also had more mental health symptoms before treatment, so the study can't say whether HRT helped or harmed. Whether HRT is right for you is something to decide with a clinician (Zühlsdorff et al., 2026).

Can a blood test predict dementia risk after menopause?

Not yet. In the UK Biobank, higher scores on a menopause signature made of 16 proteins were linked to a 15% higher risk of Alzheimer's dementia. The team aims to develop a blood test (Wood Alexander et al., 2026).

References

1. Zühlsdorff K, Langley C, Bethlehem R, Warrier V, Romero Garcia R, Sahakian BJ. Emotional and cognitive effects of menopause and hormone replacement therapy. Psychological Medicine. 2026;56:e24. https://doi.org/10.1017/S0033291725102845 · University of Cambridge summary
2. Wood Alexander M, Rabin JS, Caunca M, et al. Blood proteomics of menopause map to brain aging and dementia risk. Nature Medicine. 2026. https://doi.org/10.1038/s41591-026-04648-4 · UCSF summary
3. Keye C. Interpreting cross-sectional comparisons of menopausal status and brain outcomes. Psychological Medicine. 2026;56:e95. https://doi.org/10.1017/S0033291726104073
4. Zühlsdorff K, et al. Response to interpreting cross-sectional comparisons of menopausal status and brain outcomes. Psychological Medicine. 2026;56:e226. https://doi.org/10.1017/S0033291726104747
5. US Food and Drug Administration. FDA approves labeling changes to menopausal hormone therapy products. Press announcement, February 12, 2026. https://www.fda.gov/news-events/press-announcements/fda-approves-labeling-changes-menopausal-hormone-therapy-products
6. Gordon JL, Rubinow DR, Eisenlohr-Moul TA, Xia K, Schmidt PJ, Girdler SS. Efficacy of transdermal estradiol and micronized progesterone in the prevention of depressive symptoms in the menopause transition: a randomized clinical trial. JAMA Psychiatry. 2018;75(2):149–157. https://doi.org/10.1001/jamapsychiatry.2017.3998
7. van ’t Hof SR, Bos MGN, Straathof M, Crone EA, Hoekzema EA. Puberty, pregnancy, and menopause show shared and distinct structural changes across the lifespan. Nature Communications. 2026;17:9360. https://doi.org/10.1038/s41467-026-76755-2
8. Wood Alexander M, Rabin JS, Caunca M, et al. Blood proteomics of menopause map to brain aging and dementia risk. medRxiv preprint, 2026. https://doi.org/10.64898/2026.02.09.26345907 (used only for the group sizes and protein counts)