1. It is real, it is common, and it is not what people fear it is
Difficulty concentrating, losing words mid-sentence, forgetting why you walked into a room, misplacing things. In the forties and fifties this is one of the most common complaints there is, and it is frequently the most distressing — not because of the symptom itself but because of what people privately think it means.
The reassurance is specific rather than general, which is what makes it worth having. A 2026 study of 14,234 women aged 45–55 compared what women reported against what they scored on eight objective cognitive tasks. Perimenopausal women were substantially more likely to report cognitive symptoms than premenopausal women. Objective performance differed minimally between the groups — and perimenopausal participants were, if anything, marginally more accurate.
Consistent with that, when women reporting brain fog undergo full neuropsychological testing, around 87–89% score in the normal or near-normal range.
Two things follow, and both matter.
It is not early dementia. Dementia in midlife is rare, and the pattern here does not resemble it: measurable performance is largely intact, and for many women the fog lifts post-menopause as the brain adapts to a lower-oestrogen state. That is a genuinely different trajectory from a neurodegenerative one.
And it is not imaginary. A symptom that does not show up on a test is still a symptom. Difficulty retrieving a word under time pressure in a meeting is not the same task as a word-list test in a quiet room, and the tests are not sensitive to everything a working brain does. The correct reading is this is real and it is not the thing you are afraid of, not your tests are normal so nothing is wrong.
2. What is actually happening
Menopause is a neurological transition as much as an ovarian one. Oestrogen is not only a reproductive hormone: it helps regulate glucose uptake in the brain, supports cerebral blood flow, and interacts with the acetylcholine system that memory depends on.
Through perimenopause, oestrogen does not decline smoothly. It fluctuates violently before it falls, which is why symptoms are often worse in the transition than after it, and why a single blood test is close to useless for characterising where someone is. PET imaging work has shown measurable shifts in brain glucose metabolism across the transition, with partial recovery afterwards.
Graded Moderate: the mechanism is well supported and the imaging is real. What is not established is how much of the reported fog is caused by that directly, and how much by the things that travel with it — broken sleep, night sweats, low mood, and the ordinary cognitive load of being 48 with a job and ageing parents. Those are hard to separate and nobody has separated them.
The diagnostic blind spot worth knowing about. Because hormone levels swing day to day, a single FSH or oestradiol result frequently looks unremarkable in someone who is plainly symptomatic. Perimenopause is a clinical diagnosis made on symptoms and age, not a blood test — and women whose bloods came back “normal” are sometimes told the problem is anxiety or depression instead, and offered an antidepressant.
The careful version of that complaint, because the loud version does harm. Low mood in midlife is common, real, and sometimes genuinely depression that needs treating — and an antidepressant is not a wrong answer to it. What is wrong is reaching a psychological explanation because a hormone level looked unremarkable, when the level was never going to settle the question. Both can be true at once, and the useful thing to say in the room is not “this is not depression” but “my symptoms started with the transition and I would like that considered as well”.
3. Hormone therapy, and what the evidence does and does not support
This is where most of the disagreement lives, and the honest position is narrower than either side of the argument usually allows.
The largest synthesis found no overall effect. A 2024 meta-analysis pooled 34 randomised trials, 14,914 treated and 12,679 placebo participants. Menopausal hormone therapy had no overall effect on cognitive domain scores. Two subgroup findings were positive: oestrogen-only therapy after surgical menopause improved global cognition, and therapy initiated in midlife or close to menopause onset was associated with better verbal memory. Late-life initiation showed nothing.
Those subgroups are worth knowing and worth treating as what they are. The verbal memory result had a confidence interval running from 0.014 to 0.774 — statistically significant, and barely. Subgroups generate hypotheses; they are not the same currency as a primary endpoint, and this site applies that rule to findings it likes as well as ones it does not.
The professional position is explicit: hormone therapy is not recommended to improve cognition. That is The Menopause Society's stated view, and it is the single most useful sentence on this subject. Anyone offering MHT as a treatment for brain fog is ahead of the evidence, and if they are selling it, that is worth noticing.
What is well supported is different, and it may matter more than it sounds. MHT is effective for vasomotor symptoms and for the sleep disruption they cause. Broken sleep is a direct and well-evidenced cause of poor concentration and word-finding difficulty. So a woman whose night sweats stop and who sleeps through for the first time in two years may well think more clearly — by a route that runs through sleep rather than through any direct effect on cognition. That is a legitimate reason to consider treatment. It is not the same claim as oestrogen protecting the brain, and blurring the two is how the subject goes wrong.
Timing, and why the older trials read badly. The Women's Health Initiative Memory Study found harm — higher dementia risk — in women started on hormone therapy in their sixties and seventies, well after menopause. The timing hypothesis holds that the same treatment begun near the transition behaves differently.
For cognition specifically, that has been tested, and the answer is neither of the things people fear or hope.
KEEPS-Cog and ELITE both randomised women in early postmenopause and both came out neutral. The KEEPS continuation, reporting in 2024, found no effect roughly a decade on. So on cognitive performance in women starting near the transition, the randomised answer is no benefit and no harm — which is a real answer, and a more useful one than “unresolved”.
“Unresolved” is the fair word for dementia incidence, which is a different and much longer question. There the observational cohorts conflict, and a systematic review in Lancet Healthy Longevity in December 2025 covers the state of it.
An earlier version of this page said the timing hypothesis had never been tested by a trial designed to answer it. That is right for dementia incidence and wrong for cognition, and the distinction matters: it means a woman starting hormone therapy near menopause is not taking a cognitive risk, and is not buying a cognitive benefit either.
And the UK anchor, which is what a GP is actually working to. NICE NG23, updated in 2024, says do not offer hormone therapy to prevent dementia, and describes the evidence on cognitive symptoms as uncertain. That is consistent with everything above and it is the sentence that will be quoted back to you in a consultation.
And the labelling has moved. US regulators revised the boxed warnings on menopausal hormone therapy in February 2026. Anyone reading older material should check the current product information rather than assume the position is where it was.
What this chapter is and is not grading. Everything above is about cognition. It is not a verdict on hormone therapy overall — the case for it in symptomatic women, and the risk-benefit balance under 60 or within ten years of the last period, is a larger question about hot flushes, bone, mood and quality of life that this site does not currently cover anywhere. Do not read a Contested grade on cognition as a Contested grade on hormone therapy. They are different questions and this page only answers one.
Graded Contested for cognition specifically, and that grade is doing real work: it means competent people read the same trials differently, and it should not be read as either endorsement or dismissal. MHT for symptoms is a decision between a woman and her prescriber, and needs no cognitive justification. MHT for cognition is not supported by any guideline.
4. What is worth doing while it lasts
Ordered by how well it is evidenced, which is not the order these usually appear in.
| What | Why it earns its place |
|---|---|
| Treat the sleep, aggressively | The best-evidenced route to clearer thinking here. That means treating vasomotor symptoms if they are what is waking you, and CBT for insomnia if it is not — which outperforms hypnotics and is unglamorous. See the sleep chapter |
| Resistance training, not only cardio | Bone and muscle are being lost fastest in exactly this window, and the metabolic shift — visceral fat, insulin resistance, a worsening lipid profile — is the same cluster the dementia chapter is built on |
| Blood pressure, lipids and HbA1c, checked in this decade | Midlife is when the vascular contribution to later cognition is decided. The fifties is the highest-yield decade for it |
| A Mediterranean or MIND-style pattern | Reasonable, well tolerated, and evidenced for cardiovascular outcomes more strongly than for cognition specifically |
| Get hearing tested if there is any doubt | The single largest modifiable factor in the Lancet Commission, and under-recognised in women |
5. Two things this chapter is not doing
It is not offering off-label stimulants. There is current clinical interest in atomoxetine and lisdexamfetamine for executive dysfunction in this group. The evidence is two small trials of roughly thirty women each — which is worth stating, because a named drug with its evidence attached is far less likely to be read as a recommendation than a named drug without it. They are not licensed for this indication, and a public reference page is the wrong place to give anyone the idea of asking for them by name. It is on the page only because it is circulating and someone should say what its status is.
And it is not telling you this is all menopause. Thyroid disease, iron deficiency, B12 deficiency, sleep apnoea, depression and alcohol all present as brain fog in this age group, all are common, and all are treatable. A menopause label that stops anyone looking is doing harm of its own — the same argument the fibromyalgia topic makes about persistent symptoms. Ask for the basic bloods.