Call a Friend
Create
For decades, women have heard conflicting messages about hormone therapy and the brain. A major new study suggests there may be a reason: the relationship between menopausal hormone therapy and dementia could differ according to when treatment begins, how menopause occurs and perhaps even a woman's genetics. But the findings do not mean women should take hormones to prevent dementia.
For women trying to understand menopause, few subjects have produced more confusing messages than hormone therapy.
One generation was told that replacing estrogen after menopause might protect the heart, bones and perhaps even the brain.
Then came studies suggesting that hormone therapy could increase serious health risks.
Later came another change in thinking: perhaps the risks and benefits depended on a woman's age, how long ago menopause occurred, which hormones she took and why she was taking them.
And now dementia has entered the discussion again.
A large new study involving more than 183,000 postmenopausal women has found that women who had used hormone replacement therapy—or HRT—for at least a year had a modestly lower rate of dementia during long-term follow-up.
But the really interesting finding was not simply that HRT users appeared to do better.
It was that some groups of women appeared to do considerably better than others.
Which raises an intriguing possibility.
Perhaps asking:
“Does HRT protect the brain?”
has always been too simple a question.
First, What Did the New Study Actually Find?
Researchers from the Universities of Exeter, East Anglia and Oxford analyzed data from 183,450 postmenopausal women in the UK Biobank.
The women were followed for an average of about 13.3 years, during which 3,948 were diagnosed with dementia.
Overall, HRT use was associated with about a 10% lower risk of all-cause dementia.
For Alzheimer's disease specifically, the association was somewhat stronger: HRT users had about a 16% lower associated risk.
But there was no statistically significant reduction in non-Alzheimer's dementias overall.
Already, that tells us something interesting.
Dementia is not a single disease.
And whatever relationship exists between hormones and brain aging may not affect every form of dementia in the same way.
But then the researchers looked more closely.
That is where the study became considerably more intriguing.
Surgical Menopause Stood Out
Among women who had undergone what the study classified as surgical menopause, HRT use was associated with about a 26% lower rate of dementia.
Among women who had experienced natural menopause, the corresponding association with all-cause dementia was much smaller and did not reach conventional statistical significance.
Why might that matter?
Natural menopause is usually a gradual biological transition.
When both ovaries are surgically removed before natural menopause, however, estrogen levels can fall much more abruptly.
Previous research has linked early loss of ovarian function with adverse long-term health consequences, which is one reason hormone therapy is treated differently in women experiencing premature or early menopause.
Current clinical guidance already recognizes that women who lose ovarian function unusually early are not necessarily comparable with women undergoing menopause at the usual age. ACOG, for example, notes that hormone therapy is commonly recommended in early menopause because of longer-term health risks associated with premature estrogen loss.
But there is an important wrinkle in the new UK Biobank study.
Its “surgical menopause” category included women who had undergone either bilateral removal of the ovaries or hysterectomy, even though hysterectomy alone does not necessarily cause the abrupt estrogen loss produced by removal of both ovaries.
The authors acknowledge this distinction.
That makes the result fascinating—but not quite as clean as the headline number might suggest.
Then There Was Timing
Perhaps the most interesting result concerns when HRT was started.
The strongest associations with lower dementia risk appeared among women who began treatment between approximately 46 and 56 years of age.
Women who began considerably later did not show the same pattern.
This lends support to an idea researchers have discussed for years:
The “window of opportunity” hypothesis.
The theory is that estrogen may affect relatively healthy brain tissue differently from brain tissue in which age-related vascular or neurodegenerative changes are already developing.
In simple terms:
The same hormone given at 50 may not have the same biological effects as the same hormone first given at 70.
That could help explain one of the great apparent contradictions in menopause research.
The Study That Changed Everything
In the 1990s, many women were prescribed menopausal hormones partly because observational research suggested broad long-term health benefits.
Then the Women's Health Initiative changed medical practice dramatically.
Its memory substudy—known as WHIMS—randomly assigned women aged 65 years and older to hormone treatment or placebo.
Among women receiving conjugated equine estrogen plus medroxyprogesterone acetate, the trial found an increased incidence of probable dementia, rather than protection.
A separate estrogen-only arm involving women who had undergone hysterectomy also failed to demonstrate dementia prevention in women aged 65 to 79. When dementia and mild cognitive impairment were considered together, risk was increased.
Those were randomized controlled trials—the strongest conventional design for determining whether a treatment causes an outcome.
So how can the new study apparently point in the opposite direction?
One answer may be:
The women weren't the same.
The WHIMS participants were at least 65 when studied.
The new UK analysis suggests that associations differed considerably according to the age at which women originally began HRT.
But that explanation remains a hypothesis rather than a settled answer.
Observational Studies and Randomized Trials Are Not the Same Thing
This distinction matters enormously.
The UK Biobank study did not randomly assign 183,450 women to take HRT or a placebo.
It observed what happened to women who had already made different treatment choices.
That means the two groups may differ in ways researchers cannot completely measure.
Women who use hormone therapy may differ from women who don't in:
education;
income;
health awareness;
access to medical care;
exercise;
diet;
cardiovascular health;
and other behaviors that might themselves influence dementia risk.
Researchers statistically adjusted for many of these differences.
But statistics cannot guarantee that every difference has disappeared.
The authors explicitly identify this possible “healthy-user bias” as an important limitation and caution that residual confounding might explain at least part of the modest overall association they observed.
This is why:
“HRT users had less dementia”
does not automatically mean:
“HRT prevented dementia.”
Association is a clue.
Causation requires stronger evidence.
And “HRT” Isn't Really One Treatment
There is another surprisingly important complication.
We often speak of HRT as though it were a single drug.
It isn't.
Menopausal hormone therapy can vary by:
the estrogen used;
whether a progestogen is included;
the type of progestogen;
dose;
duration;
whether medication is taken orally or through the skin;
and whether treatment is systemic or local.
A woman who has a uterus generally needs appropriate endometrial protection if taking systemic estrogen, whereas a woman who has undergone a total hysterectomy may use estrogen alone.
Those are biologically different treatments.
Current menopause guidance also recognizes that route matters for some risks. For example, NICE notes differing venous-thromboembolism risks between oral and transdermal HRT.
Unfortunately, the UK Biobank study could not adequately distinguish estrogen-only from combined therapy, nor properly analyze dose and route of administration.
The authors describe this as an important limitation.
So even if some hormone regimens eventually prove beneficial for particular women's brains, this study cannot tell us precisely which regimen is responsible.
What About the Alzheimer's Gene?
There is another intriguing part of the study.
The researchers examined APOE ε4, the most important common genetic risk factor for late-onset Alzheimer's disease.
The association between HRT and lower dementia risk appeared stronger among women carrying at least one ε4 allele.
That sounds potentially important.
But it requires considerable caution.
When the researchers formally tested whether APOE ε4 significantly modified the effect of HRT across the whole study, the interaction was not statistically significant.
The authors themselves say the APOE result should therefore be interpreted cautiously.
In practical terms, this study does not establish that APOE ε4 carriers should take HRT to prevent Alzheimer's disease.
It does suggest an interesting research question worthy of future trials.
Those are very different conclusions.
Estrogen and the Brain Do Have a Relationship
None of this means the hypothesis is biologically implausible.
Estrogen receptors exist throughout the brain.
Estrogen participates in processes involving synaptic function, glucose metabolism, blood vessels, inflammation and neurotransmission.
The authors note that menopause therefore represents a substantial endocrine transition occurring during the decades in which the long preclinical processes leading to dementia may also begin.
That does not mean menopause causes dementia.
Nor does it mean replacing estrogen prevents it.
But it helps explain why researchers keep returning to the question.
Hormones and brain aging are not unrelated subjects.
Another Large Study Adds to the Uncertainty
If the story ended with the new UK Biobank paper, it would be tempting to conclude that researchers had finally solved the puzzle.
They haven't.
Recent data from the long-running Nurses' Health Study provide an important counterweight.
Researchers followed more than 14,000 women and found that a longer natural reproductive lifespan was associated with better cognitive trajectories.
But hormone therapy use—even when begun relatively close to menopause—was not associated with better global cognitive performance.
The Menopause Society explicitly noted that these findings support existing recommendations against using hormone therapy for dementia prevention.
This is precisely why medical science can sometimes look frustratingly untidy.
One large observational study produces an encouraging association.
Another does not.
Older randomized trials produce different results again.
Rather than pretending the disagreement doesn't exist, good science asks:
Why?
Perhaps We Have Been Looking for One Answer Where Several Exist
The new UK research raises an increasingly familiar idea in medicine:
The average result may conceal important differences between individuals.
Consider two women.
One experiences natural menopause at 52 and has no troublesome symptoms.
Another loses ovarian function surgically at 42.
Calling both simply “postmenopausal women” may hide biologically meaningful differences.
Now add:
age;
genetics;
cardiovascular health;
years of natural estrogen exposure;
type of hormone therapy;
dose;
route;
and timing.
Suddenly the simple question—
“Is HRT good or bad?”
begins to look almost impossible to answer.
A more useful question might eventually be:
Which hormone treatment, for which woman, for which reason, beginning when?
That is the direction sometimes described as precision medicine.
And it may ultimately prove particularly relevant to menopause.
So Should Women Take HRT to Prevent Dementia?
Based on current evidence:
No medical guideline supports doing that.
NICE explicitly states that neither combined nor estrogen-only HRT should be offered for the purpose of preventing dementia.
The Menopause Society similarly states that hormone therapy is not recommended to improve cognition in naturally menopausal women and continues to support recommendations against its use specifically for dementia prevention.
That does not mean hormone therapy is inappropriate.
Quite the opposite.
Systemic hormone therapy remains one of the most effective treatments for troublesome menopausal hot flashes and night sweats, and it has other established indications. For many healthy women who begin treatment near menopause, the balance of benefits and risks may be favorable when therapy is appropriately prescribed.
But:
treating menopausal symptoms
and
taking hormones to prevent Alzheimer's disease
are not the same medical decision.
The new research should not blur that distinction.
What the Study Really Changes
Perhaps the significance of this study is not that it tells women what medication to take.
It doesn't.
Its contribution may be subtler.
It challenges the idea that decades of apparently contradictory hormone research can necessarily be reduced to one universal answer.
The study suggests that menopause history matters.
Timing may matter.
Natural estrogen exposure may matter.
Genetics might matter.
The particular hormones probably matter.
And the reason a woman is taking them certainly matters.
What we still lack is the kind of randomized evidence required to prove whether carefully selected younger women actually experience a lower dementia risk because of hormone treatment.
The authors themselves conclude that randomized trials are needed before their findings can guide HRT prescribing specifically for dementia-risk reduction.
That may be less dramatic than announcing that HRT prevents Alzheimer's.
But scientifically, it is much more interesting.
Lydia's™ Take
Women have been asked to live through several generations of changing advice about menopause.
Take hormones.
Don't take hormones.
Hormones are dangerous.
Hormones were unfairly demonized.
Hormones may protect the brain.
Hormones may harm the brain.
No wonder the subject can feel bewildering.
The emerging lesson may be that medicine was asking a question that was too simple.
Perhaps HRT is neither universally “good” nor universally “bad.”
Perhaps the woman matters.
Her age.
Her symptoms.
How menopause occurred.
Her medical history.
Which hormones she takes.
When she begins them.
And perhaps one day, her genetic profile as well.
The new study does not tell women to take hormones to prevent dementia.
Nor does it overturn the randomized trials that came before it.
What it does is remind us that averages can conceal important differences between people.
And that may ultimately be the more important development.
The future of menopause medicine may not lie in finding one answer for every woman.
It may lie in getting better at asking:
What is the right answer for this woman?
Medical Disclaimer
This article is for general informational and educational purposes only and does not constitute medical advice, diagnosis or treatment.
Hormone therapy has established benefits and potential risks that vary according to a woman's age, menopause history, medical history, symptoms, treatment formulation, dose and route of administration. The research discussed here does not establish that menopausal hormone therapy prevents dementia, and current clinical guidelines do not recommend hormone therapy specifically for dementia prevention.
Women considering starting, stopping or changing menopausal hormone therapy should discuss their individual circumstances with a qualified healthcare professional. Medication should not be started, discontinued or altered on the basis of this article.
Research & Trusted Sources
Squires S, Saleh RNM, Pilling LC, et al. “Hormone replacement therapy and dementia risk among postmenopausal women: Identifying responsive subgroups in the UK Biobank.” Alzheimer's & Dementia. 2026;22:e71679. DOI: 10.1002/alz.71679.
The principal study behind this article. It followed 183,450 postmenopausal women and found HRT use was associated with a modestly lower rate of dementia overall, with larger associations in several subgroups. The authors emphasize that the study is observational and cannot establish causation.
Shumaker SA et al. Women's Health Initiative Memory Study — JAMA.
The landmark randomized study involving women aged 65 and older found that conjugated equine estrogen plus medroxyprogesterone acetate did not prevent cognitive decline and increased probable dementia incidence.
Shumaker SA et al. WHIMS estrogen-alone trial — JAMA.
In women aged 65 to 79 who had undergone hysterectomy, conjugated equine estrogen alone did not reduce dementia or mild cognitive impairment; combined dementia/MCI outcomes were increased.
National Institute for Health and Care Excellence (NICE). Menopause: Identification and Management, current guidance updated through 2026.
NICE recommends HRT for appropriate menopausal symptoms but specifically advises against offering either combined or estrogen-only HRT for the purpose of dementia prevention.
The Menopause Society — Hormone Therapy and Cognitive Health resources.
The Society regards hormone therapy as highly effective for appropriate menopausal symptoms but does not recommend it specifically for preventing cognitive decline or dementia in naturally menopausal women.
Nurses' Health Study — reproductive span, menopausal hormone therapy and cognitive decline.
Recent prospective data involving more than 14,000 women found an association between longer reproductive lifespan and better cognitive trajectories but did not find cognitive benefit associated with menopausal hormone therapy initiated close to menopause.