Perimenopause and sleep, what changes and what the research shows

Somewhere in her mid-to-late forties, a woman who has slept reasonably well for decades starts waking at three in the morning for no reason she can name. She is not anxious about anything in particular. The room is not too warm. She simply surfaces, fully alert, and lies there doing arithmetic about how many hours are left before the alarm. This pattern is common enough, and specific enough, to have a research literature behind it, though not as large a one as the scale of the complaint would suggest. Much of what does exist has only been published in the past two decades, which is one reason so many women describe reaching this stage of life with almost no advance warning of what their nights might start to look like.

Sleep changes are one of the most commonly reported features of the transition

Sleep disruption is among the most frequently reported features of the menopause transition. Women's Health Concern, the patient information service of the British Menopause Society, puts the figure at around 40 to 56 per cent of women affected, spanning difficulty falling asleep, frequent waking, early waking and sleep that does not feel recuperative (Women's Health Concern, 2026). The same guidance distinguishes ordinary disrupted sleep from chronic insomnia, defined clinically as difficulty falling or staying asleep, or early waking, at least three nights a week for three months or more, alongside daytime tiredness, low mood or poor concentration. The distinction matters because the two call for different responses, and conflating them is one of the more common ways this subject gets flattened into a single, vague complaint.

The NHS's own patient guidance names sleep disruption directly among menopause symptoms, and is candid that it can make daily life "irritable, stressed and anxious," while also noting sleep problems often worsen alongside night sweats (NHS, 2026). The advice offered is to seek an early conversation with a GP, on the grounds that timely support can reduce the knock-on effects on health, relationships and work. Nothing in this article is a substitute for that conversation; the aim here is to describe what is known, not to set out a course of action for any individual reader.

Falling oestrogen disrupts sleep by more than one route

The mechanisms are not fully mapped, but the broad shape of the evidence is reasonably consistent. Falling oestrogen appears to affect sleep both directly and indirectly, the indirect routes running through hot flushes and night sweats that interrupt sleep architecture, through mood changes that make settling harder, and through joint aches and bladder changes that add their own separate interruptions (Women's Health Concern, 2026). Ageing, stress and pre-existing sleep disorders, including restless legs syndrome and sleep apnoea, both of which become more common from menopause onwards, layer on top of the hormonal picture rather than replacing it.

A smaller but suggestive strand of laboratory research has looked at what happens to the sleeping brain itself. One 2024 study compared the electroencephalogram, the recorded electrical activity of the brain during sleep, of eight postmenopausal women against twelve younger women in the first half of their cycle. It found lower power in the delta and sigma frequency bands associated with deep sleep, and a pattern the authors described as a dampened circadian variation in sleep microstructure, present even though the postmenopausal participants were healthy sleepers who rated their own sleep quality as good (Pérez-Medina-Carballo et al., 2024). The sample is small enough that it should be read as a signal rather than a settled finding, but it points towards something structural changing in how postmenopausal sleep is organised at the level of brain rhythms, distinct from the more familiar disruption caused by a hot flush waking someone directly.

A large US cohort shows how sleep problems track the transition, not the calendar

The single most substantial piece of evidence on this subject comes from the Study of Women's Health Across the Nation, a US cohort that followed 3,045 women aged 42 to 52 at seven community sites. Kravitz and colleagues found that at baseline roughly a third of participants already reported sleep difficulties, with waking repeatedly during the night the most prevalent single complaint, and that self-reported sleep disturbance tracked the menopausal transition itself, symptom burden, bleeding-defined stage and hormone levels together, rather than tracking age alone (Kravitz et al., 2008). That distinction is one of the more useful findings in the whole literature: two women of identical age, one early in the transition and one well past it, can have very different sleep experiences, which is part of why generic advice pitched at "women over fifty" tends to land badly.

A separate, smaller UK dataset adds texture rather than scale. Reisel and colleagues surveyed 978 women attending a specialist menopause clinic and found fatigue the single most commonly reported symptom, at 96 per cent, ahead of memory difficulties at 93 per cent and concentration problems at 91 per cent, with hot flushes and night sweats ranking much lower on the list (Reisel et al., 2024). This is a clinic sample rather than a community one, so it should not be read as representative of all women in the transition, but it is a useful corrective to the assumption that vasomotor symptoms, the flushes and sweats, are always the dominant complaint. Fatigue of the kind that follows broken sleep night after night appears, in this sample at least, to be more prominent than the symptom most public conversation still centres on.

Cognitive behavioural approaches have more evidence behind them than the popular conversation suggests

Current UK clinical guidance shows a shift towards taking non-hormonal, psychological approaches seriously alongside hormonal ones. The National Institute for Health and Care Excellence's updated 2024 guideline advises menopause-specific cognitive behavioural therapy as an option for people with sleep problems, such as night-time waking, occurring alongside vasomotor symptoms, either in addition to other management options including hormone replacement therapy, or as an alternative for people who cannot or would prefer not to use it (National Institute for Health and Care Excellence, 2024). Women's Health Concern's own guidance describes cognitive behavioural therapy for insomnia as the most effective treatment for chronic insomnia specifically, deliverable in person, in groups or online over about six weeks, and separately notes that hormone replacement therapy can improve sleep both directly, through its effect on hormone levels, and indirectly, by reducing the flushes and sweats that interrupt it (Women's Health Concern, 2026).

None of this is the Observatory telling anyone what to do; it does not diagnose, treat or advise on medication of any kind. Where sleep during perimenopause is a live concern, the current guidance is clear that it is worth raising directly with a GP or a menopause specialist, who can discuss the full range of options, hormonal and non-hormonal, against an individual's own history. What the research does support is a simple and slightly overlooked point: broken sleep at this stage is common, has identifiable and partly understood mechanisms, and is not something a woman is expected to simply absorb without support. Employer-side data shows the same reality from a different angle: in a 2023 survey of UK working women aged 40 to 60, among the 1,593 who had experienced menopause symptoms, 36 per cent said the ability to have a late start after a disturbed night would help them manage their symptoms at work, one of the more practical adjustments named in the survey (Chartered Institute of Personnel and Development, 2023).

What the Observatory observes

The evidence above describes averages and mechanisms. It does not describe any individual woman's night, which is exactly why the Observatory's method is built around noticing a pattern over time rather than accepting either a single bad night or a single good one as the whole story. Members are invited to record simple observations, when sleep was disturbed, what preceded it, how the following day felt, across weeks rather than snapshots, and to try small, optional adjustments and note what changed, entirely at their own pace and never as a substitute for clinical advice.

This is also where the case for a longer view becomes concrete rather than abstract. A single bad night tells a woman very little; a month of nights, read alongside her own notes on stress, activity and context, starts to show which factors seem to travel with her sleep and which do not. Contribution of that record to the Observatory's wider, anonymised dataset is optional and separately consented, and nothing about it constitutes a result or a promise of improvement. For a broader look at why duration alone is a poor measure of rest, see sleep is more than duration; for the distinction between a working profile and a clinical finding, see the difference between a profile and a diagnosis.

Waking at three in the morning for no obvious reason is not, on this evidence, a personal failing or a sign that something has gone irreversibly wrong. It is a documented feature of a transition that current research is still working to fully explain. Observing how it moves across weeks and months, rather than judging any single night in isolation, is a modest but genuinely useful place to start.