Perimenopause Brain Fog: What Helps and What Doesn’t

Cognitive complaints during the menopause transition are real, measurable and — for most women — time-limited. In the SWAN cohort of 2,362 women followed for four years, processing speed and verbal memory stopped improving during perimenopause while premenopausal and postmenopausal women kept improving, and performance rebounded afterwards. Average performance stayed within normal limits; roughly 11–13% of women showed clinically significant impairment. Menopausal hormone therapy is not recommended at any age to treat cognitive concerns, and no supplement has been shown to help this group. The things with evidence behind them are sleep, vasomotor symptom control and mood treatment.

Written by the Fog & Clarity editorial team · Published 11 September 2026 · Last updated 11 September 2026

This page carries no affiliate link and never will. Pages about a population defined by a medical context are permanently unmonetised here, because a page that recommends a product to a woman who needs a thyroid panel is doing something worse than being wrong.

What actually changes

Three things change at once during the transition, and they are usually tangled together.

Estradiol stops being stable. Perimenopause is not a slow decline; it is erratic. Levels swing widely from cycle to cycle before settling low. Estrogen receptors are densely expressed in the hippocampus and prefrontal cortex, which is the mechanistic reason people expect cognitive effects — but mechanism is not effect, and as the hormone-therapy trials below show, putting the hormone back does not restore the cognition.

Sleep gets worse. Night sweats fragment sleep; so does the insomnia that often arrives independently. Fragmented sleep degrades memory encoding, not just recall. In a controlled experiment, 28 healthy adults kept awake for about 35 hours before a learning session showed a 19% deficit on a later recognition test, with significantly decreased activation in bilateral posterior hippocampus during encoding.¹ That study was in 18–30-year-olds and says nothing quantitative about a 49-year-old with night sweats — but it identifies the mechanism that a bad night acts through.

Mood shifts. Depression and anxiety both peak in incidence around the transition, and impaired concentration is a diagnostic feature of depression rather than a side effect of it. The 2022 International Menopause Society white paper is explicit that sleep difficulties and low mood are associated with cognitive difficulties at menopause, and that frequent vasomotor symptoms strongly relate to memory difficulties.²

That last point matters for what you do about it. If the fog is downstream of hot flashes waking you five times a night, treating the hot flashes is a cognitive intervention.

What the evidence says for this group — and only this group

The best data come from the Study of Women’s Health Across the Nation. Greendale and colleagues followed 2,362 women for four years across menopause stages, testing processing speed (Symbol Digit Modalities), verbal memory (East Boston Memory Test) and working memory (Digit Span Backward).³ The finding is subtler than “memory declines,” and the subtlety is the point.

Study	Population	Design	Duration	Outcome
SWAN (Greendale 2009)	2,362 women across the transition	Longitudinal cohort, repeat testing	4 years	Late perimenopausal women did not improve on processing speed (p = 0.2) while pre-, early peri- and postmenopausal women did (p ≤ 0.0008); verbal memory plateaued during peri and climbed again after
KEEPS-Cog (Gleason 2015)	693 recently postmenopausal women, mean age 52.6	RCT: oral CEE vs transdermal estradiol vs placebo	Up to 4 years	No treatment-related benefit on any cognitive outcome; oral CEE improved depression and anxiety scores, transdermal did not
KEEPS Continuation (Gleason 2024)	275 of the original participants	Observational follow-up	~10 years post-trial	No long-term cognitive benefit and no long-term cognitive harm
WHIMS (Shumaker 2003)	4,381 women over 65, mean age 71	RCT: CEE 0.625 mg + MPA 2.5 mg vs placebo	Mean 4.2 years	Doubled risk of probable dementia (HR 2.05, 95% CI 1.21–3.48)

Read those four rows together and a clear picture emerges. The measurable dip is confined to the transition itself and largely resolves. Hormone therapy started in early menopause does not improve cognition, and ten years on it has neither helped nor hurt. Hormone therapy started in women in their seventies was associated with doubled dementia risk.⁴⁵⁶

The IMS white paper puts the clinical bottom line plainly: hormone therapy is not recommended at any age to treat cognitive concerns at menopause or to prevent cognitive decline or dementia.² That is not an argument against hormone therapy — it remains a first-line treatment for vasomotor symptoms, and if it stops the night sweats that are wrecking your sleep, the cognitive benefit arrives by that indirect route. It is an argument against taking it for your memory.

⚠️ The generalisation error this field is built on

Almost every supplement marketed for “menopause brain fog” cites research done in adults over 65, often with mild cognitive impairment or dementia. A result in a 74-year-old with MCI tells you nothing about a 48-year-old in late perimenopause. The populations differ in baseline cognition, in what is physiologically happening, and in what an intervention could plausibly do. When you see a study cited on a product page, the first thing to check is not the p-value. It is the age and diagnostic status of the people in it.

The other thing worth stating: about 11–13% of women in this literature show clinically significant impairment, while the average stays within normal limits.² Both halves of that sentence are true and both get dropped by the people who quote it. “Most women are fine on testing” does not mean your experience is imaginary. “Cognitive change is documented” does not mean everyone is impaired.

What else could it be — and why it’s worth ruling out first

Perimenopause arrives at an age when several other causes of cognitive complaint become more common, and it makes an excellent hiding place for them. Before accepting “it’s just my hormones,” it is worth having a clinician consider:

Thyroid disease. Hypothyroidism produces fatigue, cold intolerance and impaired concentration, and its incidence rises with age in women.
Iron deficiency. Heavy and irregular bleeding is one of the defining features of perimenopause, and iron deficiency is associated with weakness, fatigue, difficulty concentrating and impaired cognitive function.⁷
Vitamin B12 deficiency. Correcting a genuine deficiency matters; supplementing someone who is already replete does not.⁸
Depression, anxiety and untreated sleep apnea. All three produce exactly this symptom picture, and all three are treatable.
Medication burden. Anticholinergic load from antihistamines, bladder medications and some antidepressants has better evidence for cognitive effects than most supplements do.

Our list of symptoms that should send you to a doctor rather than a search engine is at brain fog red flags. If you have already been through the panel and it came back clean, this page is the next step.

Non-supplement interventions, in order

1. Treat the vasomotor symptoms. This is the intervention with the clearest line to the complaint. Frequent hot flashes relate strongly to memory difficulties in this population, and night sweats fragment sleep.² Hormone therapy, non-hormonal prescription options and cognitive behavioural therapy for vasomotor symptoms are all clinician conversations, and all of them are better uses of an appointment than asking about a nootropic.

2. Protect sleep specifically, not generically. The goal is unbroken blocks, not hours in bed. Practically: keep the bedroom cold, treat the night sweats, get a sleep apnea assessment if you snore or wake unrefreshed, and be deliberate about caffeine timing — the evidence on late-day caffeine and sleep disruption is dose-dependent and stronger than most people assume. Our sleep debt page covers what recovers and what doesn’t.

3. Treat mood, honestly. If low mood, loss of interest or anxiety are in the picture, they are not a personality failing to be worked around; they are the most treatable thing on this list, and concentration usually improves when they are addressed.

4. Exercise. It is the intervention with the broadest evidence base in cognitive health and nobody has a commercial reason to promote it, which is why it gets a fraction of the coverage that botanicals get. We cover it at exercise and memory.

5. Reduce the load rather than raise the capacity. Write things down. Externalise. This is the same advice we would give anyone whose encoding is being degraded by poor sleep, and it works because it bypasses the broken step entirely rather than trying to repair it.

Where a supplement might be reasonable

Short answer: almost nowhere, and never as the first move. There is no compound with a demonstrated cognitive benefit in perimenopausal women specifically. That sentence is the honest state of the field, and any page that tells you otherwise is either citing a different population or citing nothing.

Three narrow exceptions, all conditional:

Correcting a documented deficiency. If bloodwork shows iron or B12 deficiency, correct it — because the deficiency is worth correcting, not because supplements do anything in people who are replete.⁷⁸ ⚠️ Iron and calcium-containing multivitamins interfere with levothyroxine absorption; separate them by four hours.

A plain daily multivitamin. The COSMOS trials are the only large, properly powered, multi-year RCTs in this whole space with a positive replicated cognitive result — and the winner was a supermarket multivitamin-mineral, not a botanical blend.⁹ The caveat is the entire point of this page: COSMOS was run in older adults, with the cognitive sub-studies in participants averaging well over 65. It is not evidence for anything in a 48-year-old. We mention it because it is the most defensible thing in the category, not because it has been shown to work here.

Creatine, with the population caveat attached. A 2023 meta-analysis of 8 RCTs in 225 healthy participants found a memory effect of SMD 0.29 — concentrated in adults aged 66–76 (SMD 0.88) and essentially absent in those aged 11–31 (SMD 0.03).¹⁰ Midlife sits between those brackets, and nobody has tested it there. The pharmacology suggests the effect depends on low baseline brain creatine or high metabolic stress, which would predict some benefit in a chronically sleep-deprived midlife adult — but that is reasoning, explicitly labelled as reasoning, not a finding.

Community reports — not trial data

Menopause forums contain a large volume of reports that a specific supplement “gave me my brain back,” and an equally large volume of reports that the same supplement did nothing. Both are consistent with a condition that fluctuates, that has a substantial placebo-responsive symptom component, and that improves on its own timescale. People who improve while taking something are far more likely to post than people who improve while taking nothing. Treat these as hypotheses about your own n-of-1, tracked for six weeks with a written log, not as evidence.

What to be sceptical of

The menopause supplement market has specific tells, and once you know them the category becomes much easier to navigate.

Botanicals borrowed from the hot-flash literature

Black cohosh, dong quai, evening primrose oil, red clover and DHEA are marketed for menopause generally, and the cognitive claim is a bolt-on. NCCIH’s assessment of the underlying evidence is unflattering across the board: a 2012 review concluded there is not enough evidence to support black cohosh for menopause symptoms, isoflavone results are inconsistent, very little research exists on dong quai or evening primrose, and DHEA’s long-term safety is unknown.¹¹ ⚠️ NCCIH also flags rare cases of liver damage, some serious, in people taking commercial black cohosh products, and notes that even short-term DHEA use may have harmful effects including liver damage.

Ginkgo, still

Ginkgo is the most persistent memory botanical and the one with the widest gap between belief and data. The 2026 Cochrane review covering 82 studies and 10,613 participants found it probably makes little or no difference in mild cognitive impairment at six months, with a modest and variable effect confined to established dementia.¹² Two enormous prevention trials — GEM (n≈3,069, 6.1 years) and GuidAge (n≈2,854, 5 years) — both failed. ⚠️ It also carries a real bleeding-risk interaction with anticoagulants and antiplatelets, and should be stopped at least two weeks before surgery.¹³ We cover the trials at two enormous trials, two failures and the interaction at ginkgo and blood thinners.

Adaptogens sold to a population that is often on thyroid medication

⚠️ Ashwagandha is heavily marketed to midlife women for stress and sleep. It has a documented interaction with thyroid hormone — it can increase thyroid hormone levels — and it carries a LiverTox likelihood score of B for liver injury, with a typical latency of two to twelve weeks after starting.¹⁴ Given how many women in this age band take levothyroxine, this combination deserves a prescriber’s involvement rather than a shopping decision. See ashwagandha and levothyroxine.

Anything promising restoration

“Restore years of memory,” “reverse brain aging,” “get your sharpness back” — this framing is what the FTC litigated over in the Prevagen case, and in December 2024 a federal court ordered the marketer to stop making the deceptive memory and cognitive-improvement claims after a jury trial. In 2019 the FTC and FDA jointly sent warning letters to three companies over roughly 15 products making claims about treating Alzheimer’s and other serious diseases.¹⁵ A supplement legally cannot claim to treat a disease; a product that does so is telling you something about its own regulatory literacy.

Brain training apps positioned as the “drug-free” option

The 2016 review in Psychological Science in the Public Interest examined the field and found that brain-training interventions improve performance on the trained tasks, with little evidence that training enhances performance on distantly related tasks or improves everyday cognitive performance.¹⁶ You get better at the game. See do brain training apps work.

What would change our mind

A randomized controlled trial recruiting women in late perimenopause specifically — not postmenopausal women, not women over 65 — with objective processing-speed and verbal-memory endpoints, running at least twelve months, and testing an intervention against placebo with adequate power. That trial does not exist for any supplement in this category. If one ran and found a genuine effect on measured cognition in this population, we would say so on this page, and we would still not put a purchase link on it.

Frequently asked questions

Does perimenopause brain fog go away?

For most women, yes. In SWAN, processing speed and verbal memory stopped improving during perimenopause and resumed improving in postmenopause, which the authors interpreted as transition-related cognitive difficulties being time-limited. That is a group average over four years, not a promise about any individual, and the IMS white paper notes difficulties persist for some women into postmenopause.

Will HRT fix my memory?

The direct evidence says no. KEEPS-Cog randomized 693 recently postmenopausal women to oral conjugated equine estrogens, transdermal estradiol or placebo for up to four years and found no treatment-related benefit on any cognitive outcome. Ten years later, the continuation study found neither long-term benefit nor long-term harm. Hormone therapy may still help your cognition indirectly by stopping the night sweats that are fragmenting your sleep — that is a real and worthwhile route, and it is a different claim.

Is this early dementia?

Transition-related cognitive change is characterised by plateaued improvement rather than progressive decline, and average performance stays within normal limits. Symptoms that are steadily worsening month over month, that other people notice before you do, or that involve getting lost in familiar places, losing words mid-sentence, or difficulty with previously routine tasks are a different pattern and warrant assessment. Our red flags page lists them.

Why does my doctor say my bloodwork is normal?

Because the transition itself does not show up on a standard panel — the panel is there to rule out the things that would. A normal result is genuinely useful information: it removes thyroid disease, anemia and B12 deficiency from the list, which narrows what is left. What it does not do is validate the symptom, which is why the conversation often feels dismissive.

Are there any supplements with real evidence in this group?

No. Not one compound has been shown to improve measured cognition in perimenopausal women specifically. Several have evidence in adults over 65, and a few of those are marketed as though the finding transfers. It does not.

Sources

These statements have not been evaluated by the Food and Drug Administration. Nothing here is intended to diagnose, treat, cure or prevent any disease.

This article is information, not medical advice. Talk to a doctor or pharmacist about your own situation, especially if you take prescription medication.