Brain Fog, Dementia, and the Hormone Replacement Question
Rahul Kapur, MD — KC Primary Care, Leawood, KS
A patient sat in my office recently and said something I hear often, usually in a lowered voice,
usually near the end of a visit: “I think something is wrong with my brain.”
She was 47. She had walked into a room three times that week and forgotten why. She had lost
a word mid-sentence during a client presentation. She had started writing everything down
because she no longer trusted herself to remember it. She had a grandmother with Alzheimer’s,
and she had convinced herself she was watching the same thing begin in herself.
She was not developing dementia. She was in perimenopause.
But her fear was not irrational, and it deserves a better answer than “it’s just your hormones,
you’ll be fine.” Here is what the evidence actually shows — including the parts that complicate
the story you may have read online.
Brain fog is real, and it is measurable
For a long time, cognitive complaints during the menopause transition were treated as stress,
poor sleep, or anxiety. That dismissal was wrong.
Roughly two-thirds of women report cognitive difficulties during the menopause transition.
Studies looking specifically at perimenopause find measurable effects on verbal learning, verbal
memory, processing speed, attention, and working memory. These are not vague complaints —
they show up on testing.
The mechanism makes biological sense. Estrogen receptors are distributed throughout the
brain, including regions central to memory and executive function. Estrogen supports how
neurons use glucose for fuel, influences neurotransmitter systems tied to memory and attention,
regulates cerebral blood flow, and has anti-inflammatory effects in the central nervous system.
Neuroimaging work from Lisa Mosconi’s group at Weill Cornell has documented changes in
brain structure, connectivity, and energy metabolism across the menopause transition.
Progesterone matters here too, and it is often the first hormone to decline as ovulatory cycles
become less frequent. Its metabolite allopregnanolone acts on GABA receptors — the brain’s
primary inhibitory system — which is part of why falling and fluctuating progesterone tracks so
closely with anxiety, irritability, and disrupted sleep in early perimenopause.
One detail is worth emphasizing, because it explains why perimenopause often feels worse than
postmenopause: the problem may be the volatility, not just the decline. In perimenopause,
hormone levels swing unpredictably from day to day. The brain adapts well to a new steady
state. It adapts poorly to a target that keeps moving.
Brain fog is not early dementia
This is the part I want women to actually hear.
Perimenopausal cognitive symptoms and early dementia are different phenomena, and they are
usually distinguishable:
- Perimenopausal brain fog looks like word-finding trouble, losing your train of thought,
walking into a room and forgetting your purpose, difficulty juggling tasks, and slower
processing. Insight is fully preserved — you know it is happening, and it frustrates you. - Early dementia looks like difficulty performing familiar tasks, getting lost in familiar
places, trouble following directions, personality change, and — critically — often reduced
awareness that anything is wrong.
The illustration that I have heard used here and use myself with my patients : forgetting the
word “fork” is perimenopause. Forgetting how to use a fork is not.
There is also reassuring longitudinal data. Cognitive symptoms during the transition are
generally mild and tend to improve after menopause. The brain changes seen on imaging
during perimenopause appear to represent a period of reorganization and adaptation, not the
beginning of a neurodegenerative slide.
That said — if symptoms are severe, progressive, or interfering with your ability to function
independently, that is a reason for formal evaluation rather than reassurance. Iron deficiency
(very common with the heavy bleeding of perimenopause, and missed if only hemoglobin is
checked rather than ferritin), thyroid disease, B12 deficiency, sleep apnea, depression, and
medication effects all mimic this picture and are all treatable.
Why women’s dementia risk is a legitimate question
About two-thirds of Alzheimer’s patients are women. That statistic is real and it is not explained
solely by women living longer.
This is why the menopause transition has drawn so much research attention. If the
neuroendocrine shift at midlife contributes to that disparity, it would be a modifiable window.
That hypothesis is scientifically reasonable. Whether hormone therapy actually acts on it is a
separate question — and one where the evidence has gotten more interesting, not more settled.
The hormone replacement question
Here is where I have to resist giving you a cleaner answer than the data supports. Four findings,
and they do not fully agree.
- The largest synthesis to date found no effect either way. A systematic review and
meta-analysis published in Lancet Healthy Longevity in December 2025, led by a UCL team,
pooled over one million participants across ten studies. It found no significant association
between menopausal hormone therapy and dementia or mild cognitive impairment. Subgroup
analyses by timing, duration, and formulation did not change that. Certainty of evidence was
rated very low for most outcomes. - The best randomized long-term data found no benefit. The KEEPS Continuation study
re-evaluated women roughly a decade after they had been randomized to four years of oral
conjugated estrogens, transdermal estradiol, or placebo, started within three years of
menopause. Cognitive performance was equivalent across all three groups. The authors’
conclusion was deliberately two-sided: this is reassuring about long-term safety, and it also
means hormone therapy should not be prescribed to preserve cognition. - But a 2026 autopsy study pointed the other direction. In August 2026, Stanford
investigators published in Neurology the first study to examine hormone therapy against the
gold standard — actual Alzheimer’s pathology in autopsied brains. Among women who had
used estrogen-only therapy, they found significantly lower amyloid plaque and neurofibrillary
tangle burden, and roughly a 35% reduction in the odds of developing Alzheimer’s disease.
Estrogen-only users also had substantially lower odds of receiving a dementia diagnosis in their
lifetime and performed better on memory testing. - Regulatory posture has shifted. In late 2025 the FDA removed the black box warning from
menopausal hormone therapy products — warnings that had included the claim of increased
dementia risk, which originated in the 2003 WHIMS analysis.
Why estrogen alone and estrogen-plus-progestogen are not the same
question
Notice what the strongest signals in both directions have in common: they are about
estrogen-only therapy.
This is not a coincidence, and it is the single most important nuance in this entire topic.
In the original WHIMS trial, the arm that showed a statistically significant increase in dementia
risk was conjugated equine estrogens combined with medroxyprogesterone acetate (MPA), a
synthetic progestin. The estrogen-only arm showed a non-significant increase that did not reach
statistical significance. The headline that circulated for two decades — “hormone therapy
causes dementia” — was largely driven by the combination arm, in women whose average age
at initiation was 72.
The Stanford autopsy data, conversely, could only draw conclusions about estrogen-only users.
There were too few women on estrogen-plus-progestogen regimens in the autopsy cohort to
say anything meaningful.
So we have relatively better evidence about estrogen alone, in both directions, and genuinely
thin evidence about the combination — which is what most women with an intact uterus actually
take.
Is micronized progesterone different from MPA?
This is the question I get from patients who have read about “bioidentical” hormones, and it
deserves a careful answer rather than an enthusiastic one.
Pharmacologically, they are clearly not the same molecule. MPA is a synthetic
17α-hydroxyprogesterone derivative compared to micronized progesterone
For cognition specifically, the comparative evidence is thin.
My honest read: I use micronized progesterone in my practice, and I think the case for it over
MPA is reasonable — mostly on breast and cardiovascular grounds, and on the general
principle of using the molecule the body actually makes. But I would be overstating the evidence
if I told you it has been shown to protect your brain, or even that it has been shown to be
cognitively safer than MPA. It has not. Anyone telling you otherwise is extrapolating from breast
cancer data and mechanism.
And to be clear about why progesterone is there at all: if you have a uterus, unopposed
estrogen causes endometrial hyperplasia and raises endometrial cancer risk. Progesterone is
not optional in that setting. It is endometrial protection, and that is a strong indication
independent of any cognitive consideration.
Timing: the critical window
The observation that keeps recurring across studies is that when therapy starts appears to
matter more than whether it starts.
Trials that enrolled women aged 65 and older — WHIMS in particular — showed harm.
Observational research on midlife initiation has more often shown neutral or favorable
associations, particularly for estrogen-only therapy. This is the basis of the “critical window” or
timing hypothesis: that estrogen may support neurons that are still metabolically healthy, but
does little or nothing for a brain in which degeneration is already underway.
The hypothesis is credible and mechanistically sensible. It has not been proven by a
randomized trial, because the trial that would prove it — randomizing women in their late forties
and following them for thirty years — has never been run and probably never will be.
Route of administration
KEEPS randomized women to oral conjugated estrogens and to transdermal estradiol, and
found no cognitive difference between them at ten years. The Stanford autopsy findings applied
only to oral estrogen, because topical users were excluded from that analysis.
There are good reasons to prefer transdermal estradiol in many women — it avoids first-pass
hepatic metabolism and carries lower venous thromboembolism risk. Those reasons are
cardiovascular and thrombotic, not cognitive. I do not have evidence that route of delivery
changes your dementia risk.
Where the evidence is strongest: early and surgical menopause
If there is one group for whom the hormone-and-brain question has a clearer answer, it is
women who lose ovarian function early.
Premature ovarian insufficiency and early menopause — and especially surgical menopause
from bilateral oophorectomy before the natural age of menopause — are associated with
increased risk of cognitive impairment. For these women, estrogen therapy until roughly the
average age of natural menopause is standard of care, and it is recommended by major
menopause societies. Analyses from the UK Biobank have found hormone therapy associated
with reduced dementia risk specifically in women with surgical menopause and in those with
shorter lifetime estrogen exposure.
This is a different clinical situation from a 52-year-old going through natural menopause on
schedule, and it should not be reasoned about the same way.
What I actually tell patients
Hormone therapy is an excellent treatment for menopausal symptoms, including the
cognitive symptoms that accompany them. It is not currently a justified prescription for
dementia prevention.
Those are different clinical decisions, and conflating them does women a disservice in both
directions — it oversells treatment to women who don’t need it, and it frightens women who
would benefit into avoiding it.
If you have significant symptoms, symptom relief is a sufficient reason to consider treatment on
its own. The evidence on long-term cognitive safety is now genuinely reassuring, which was not
something I could say with confidence ten years ago. Much of the cognitive improvement
women report on hormone therapy is likely real, and likely mediated through better sleep and
the elimination of night sweats rather than through any direct effect on neurons.
If your primary goal is preventing dementia, hormone therapy is not the intervention with the
best evidence behind it. The primary goal would be fracture prevention and cardiovascular
disease and there is good data for both
Cardiovascular disease is the dominant cause of death in women, and early-initiation HT
appears protective.
Fracture prevention was the one unambiguous WHI benefit and was ignored. Hip fracture
carries mortality comparable to breast cancer, and HT remains among the most effective
preventive interventions for it.
A brief word on testosterone
Since it comes up in nearly every conversation about this: testosterone is frequently promoted
online as the missing piece for brain fog, and the evidence does not support that.
It has a great positive effect on mood, anxiety , strength , energy , sleep and libido – per the
menopause society Its only evidence-based indication is hypoactive sexual desire disorder in
postmenopausal women, where the effect is moderate and real.
Testosterone has a legitimate place in women’s care. Treating brain fog is not currently one of
its established uses, and women deserve to hear that plainly.
What actually protects the brain
The interventions with the strongest evidence are, frustratingly, the ones nobody wants to hear
about. They also work.
- Treat the sleep, not just the symptom. Memory consolidation depends on slow-wave
sleep. Night sweats fragment it. Vasomotor symptom treatment often improves cognition
through this route alone, and that may be much of what women experience as “HRT
fixed my brain fog.” - Aerobic exercise and resistance training. Improves cerebral blood flow and BDNF,
and preserves the muscle mass and insulin sensitivity that decline sharply at midlife. - Cardiometabolic control. Blood pressure, lipids, glucose, and insulin resistance are
among the largest modifiable dementia risk factors, and midlife is when they matter
most. - MIND or Mediterranean dietary pattern.
- Rule out the mimics. Ferritin, full thyroid panel, B12, sleep apnea screening. I check
these before I attribute anything to hormones. - Hearing loss, alcohol, smoking, social isolation, and untreated depression are all
on the Lancet Commission’s modifiable risk list and all deserve attention in midlife.
The bottom line
If you are in your forties or fifties and your brain feels unreliable: you are not imagining it, and it
is almost certainly not dementia.
Estrogen and progesterone therapy may help you feel considerably better, and we now have
good evidence it will not harm your brain long-term. It is not a dementia prevention strategy, and
any clinician who sells it to you as one is ahead of the data. If you went through menopause
early or surgically, the calculus is different and more favorable — talk to someone about it.
What will move the needle on your long-term brain health is much less glamorous: sleep,
cardiometabolic risk, movement, and finding the treatable things that are hiding under the label
“brain fog.”
That is a conversation worth having with an actual physician who has time to have it.
If cognitive symptoms are affecting your work or your quality of life, we can evaluate
them properly — including the labs that commonly get skipped.
KC Primary Care 11709 Roe Ave, Suite B, Leawood, KS 66211 913-350-0586
Selected references
- Melville M, He L, Desai R, et al. Menopause hormone therapy and risk of mild cognitive
impairment or dementia: a systematic review and meta-analysis. Lancet Healthy Longev.
2025;6(12):100803. - Bruno J, Shaw J, Hosseini H, et al. Menopausal hormone therapy and Alzheimer’s
disease neuropathology. Neurology. Published online August 12, 2026. - KEEPS Continuation Study. Long-term cognitive effects of menopausal hormone
therapy. PLOS Med. 2024;21(11):e1004435. - Shumaker SA, Legault C, Kuller L, et al. Conjugated equine estrogens and incidence of
probable dementia and mild cognitive impairment in postmenopausal women (WHIMS).
JAMA. 2004;291:2947-2958. - Sherwin BB, Grigorova M. Differential effects of estrogen and micronized progesterone
or medroxyprogesterone acetate on cognition in postmenopausal women. Fertil Steril.
2011;96(2):399-403. - Wang Y, Islam RM, Bond M, Davis SR. Testosterone and pre-androgens by age and
menopausal stage at midlife. eBioMedicine. 2025;121:105972. - Davis SR, Baber R, Panay N, et al. Global Consensus Position Statement on the Use of
Testosterone Therapy for Women. J Clin Endocrinol Metab. 2019;104(10):4660-4666. - Mosconi L, Rahman A, Diaz I, et al. Menopause affects human brain structure,
connectivity, energy metabolism, and amyloid-beta deposition. Sci Rep. 2021;11:10867. - Greendale GA, Karlamangla AS, Maki PM. The menopause transition and cognition.
JAMA. 2020;323(15):1495-1496. - Rocca WA, Bower JH, Maraganore DM, et al. Increased risk of cognitive impairment or
dementia in women who underwent oophorectomy before menopause. Neurology.
2007;69(11):1074-1083.
This article is for general education and does not replace individualized medical advice.