Brain fog, mood and memory in midlife: what the evidence actually shows

You lose the word. Not an obscure word, a normal one, in the middle of a sentence you were confident about. Or you walk into a meeting having genuinely forgotten what it was for. If you have started keeping a private tally of these moments and quietly worrying about what they mean, you are in very large company, and the research is more reassuring than the worry.

What women report, and whether measurement agrees

About two thirds of women report memory complaints such as forgetfulness during the menopause transition ¹.

For a long time the standard response was that this was probably about sleep, or stress, or getting older. Then longitudinal testing was done, and it partly backed the complaints up.

SWAN tracked verbal memory, working memory and processing speed in women moving through the transition. In premenopause, scores on verbal memory and processing speed improved with repeat testing, which is the normal practice effect you expect when someone sits the same test annually. In perimenopause, that improvement disappeared ¹.

That is a subtle finding and an important one. The deficit was not a collapse in performance. It was a loss of the ability to improve, which lines up with the lived experience of feeling capable but slower and less able to absorb new things.

The other half of the finding matters just as much: the perimenopausal decrement appears to be time-limited, because improvement with practice returned in early postmenopause ¹.

What is ageing rather than menopause

Some midlife cognitive change is genuine ageing, and it starts earlier than most people assume.

SWAN researchers designed a study specifically to separate the two, using the third cognition visit as the baseline to reduce practice effects and adjusting for the transition and its symptoms. In 2,124 women with a median 6.5 years of follow-up and average baseline age of 54, two of four cognitive measures declined. Processing speed fell by an average of 0.28 points per year, about 4.9% over ten years, and delayed verbal episodic memory fell about 2% over ten years ².

Those are real, measurable within-woman declines, and they were present after adjusting for depressive, anxiety, vasomotor and sleep symptoms ².

SWAN's summary of the timing is that after menopause, which occurs on average at age 52, cognitive processing speed fell, while declines in verbal and working memory did not typically begin until after ages 58 and 61 respectively ¹.

There is also a modifiable thread running through this. Women with high blood pressure, elevated glucose and obesity, and women who reported financial hardship, experienced faster declines in processing speed ¹. Whether controlling those conditions slows cognitive decline is being tested in ongoing trials ¹.

What is not causing it

Some of the most useful findings here are negative.

During the transition itself, SWAN found that self-reported measures of sleep and hot flushes were not associated with decrements in memory, processing speed, or practice-related improvement ¹. This runs against the intuitive explanation, and it means "you are just tired" is not a sufficient account of perimenopausal brain fog.

In early postmenopause the picture changes: women with greater wakefulness and sleep fragmentation did score lower on processing speed ¹. Sleep complaints in this period are themselves multifactorial, and hot flushes account for only part of the measured wakefulness ⁸.

Mood, by contrast, showed up. Women with depressive symptoms performed less well on tests of processing speed, and women with anxiety symptoms showed smaller practice-related improvements in verbal memory ¹.

So the honest summary is that perimenopausal cognitive change is partly its own phenomenon, partly overlapping with mood, and partly the leading edge of normal ageing, and no clinician can cleanly separate those three in an individual woman.

Mood: the overlap that complicates everything

Mood in midlife is not a footnote to cognition. It is a serious clinical issue in its own right, and it is the part of this article where getting help matters most.

An expert panel convened to write guidelines on perimenopausal depression described the perimenopause, meaning the early and late transition stages plus early postmenopause, as a window of vulnerability for the development of both depressive symptoms and major depressive episodes ³.

Several points from those guidelines are worth carrying into an appointment.

Most midlife women who experience a major depressive episode during perimenopause have had a prior episode of depression ³. Your history is directly relevant, and worth mentioning even if it was decades ago.

Midlife depression presents with classic depressive symptoms, commonly combined with menopause symptoms such as vasomotor symptoms and sleep disturbance, and with psychosocial challenges ³. The guidelines say plainly that menopause symptoms complicate, co-occur and overlap with the presentation of depression ³.

Diagnosis involves identifying menopausal stage, assessing both psychiatric and menopause symptoms, considering the psychosocial factors common in midlife, working through differential diagnoses, and using validated screening instruments ³. If your appointment involves a questionnaire, that is the process working.

On treatment, the guidelines are direct: proven therapeutic options for depression, meaning antidepressants and psychotherapy, are the front-line treatments for perimenopausal depression ³. On hormones they are careful. Oestrogen therapy is not approved to treat perimenopausal depression, but there is evidence that it has antidepressant effects in perimenopausal women, particularly those with concurrent vasomotor symptoms. Data on oestrogen plus progestin are sparse and inconclusive ³.

How NICE frames the same territory

NICE draws a line between depressive symptoms and diagnosed depression, and treats them differently.

For depressive symptoms that do not meet the criteria for a diagnosis of depression and that started around the same time as other menopause symptoms, NICE says consider HRT ⁴. It also says consider CBT as an option for people with depressive symptoms in association with vasomotor symptoms ⁴.

For people suspected of having, or diagnosed with, depression, NICE says follow the menopause guideline alongside its guideline on treating and managing depression in adults ⁴.

The NHS notes that if you have low mood, depression or anxiety related to perimenopause or menopause, you may be offered antidepressants ⁶.

The distinction is not bureaucratic. It determines whether the conversation is about your HRT dose or about treating a depressive illness, and those are different conversations that sometimes need to happen at once.

Get help now if

Get urgent help if you have thoughts of harming yourself or ending your life. If you are in immediate danger, contact emergency services. If not, you can talk to someone today, free and confidentially: find help numbers where you live. This is not something to raise at your next scheduled menopause review.

Call emergency services for a sudden severe headache

Cognitive symptoms that arrive suddenly are different from ones that build over months. The NHS says to call emergency services or go to an emergency department for a headache that started suddenly and is extremely painful, or a headache with a seizure, numbness or weakness in the body or face, difficulty speaking, balancing, walking or remembering things, drowsiness or confusion, or loss of vision ⁷. Not sure of the number: find help numbers where you live.

Guidance in the US and Europe

In February 2026 the FDA approved label changes that remove probable dementia, along with cardiovascular disease and breast cancer, from the boxed warning on hormone therapy products ⁹. That changes the warning on the label. It does not make hormone therapy a dementia treatment: The Menopause Society notes that for women who start hormone therapy after 60, or more than 10 years after menopause, the balance is less favourable, partly because of a greater absolute risk of dementia ¹⁰.

In Europe, the European Menopause and Andropause Society (EMAS) lists mood among the things appropriately timed hormone therapy can help, and stresses that treatment must be individualised to each woman's age, symptoms, risk profile and preferences ¹¹.

What is still uncertain

Whether hormone therapy improves cognition is not settled. NICE explicitly says do not offer combined or oestrogen-only HRT for the purpose of dementia prevention, and notes that dementia risk might increase with combined HRT if it is started at 65 or over ⁴.

Whether treating midlife blood pressure, glucose and weight slows the cognitive decline associated with them is under investigation and not yet answered ¹.

And the relationship between measured hot flushes and memory is unresolved. SWAN's self-report analysis found no association during the transition, while the SWAN fact sheet notes that other studies using objectively measured hot flushes have found an association with memory difficulties ¹. Both findings exist. Neither settles it.

When to see a clinician

SWAN advises reporting memory changes that come on suddenly to a doctor ¹.

Seeing a GP is reasonable if brain fog is affecting your work or safety, if low mood or anxiety is not lifting, or if you cannot tell which of your symptoms is driving the others. The NHS lists mood changes, poor memory and brain fog among menopause symptoms, and notes they may feel worse if you also have sleep problems and are very tired ⁵.

Your mental health history, including old episodes and what helped then, is relevant: the perimenopausal depression guidelines note that treatment selection should be guided by previous treatment responses ³.

Writing it down helps more than it should

Brain fog is the symptom people describe worst in appointments, because the whole problem is that you cannot reliably recall the instances.

A short note at the time works better than trying to summarise: the day, what happened, and what it cost you. Three lost words in a presentation is more useful information than "my memory is terrible". The same applies to mood, where a daily one-word rating over a fortnight tells a clinician something a general impression cannot.