Why women burn out differently, and why the evidence is thin

A woman in her late forties describes the same week to three different people and gets three different explanations for what is wrong with her: she is overworked, she is peri-menopausal, she is simply not coping. None of the three is measuring anything. They are pattern-matching against whatever story is closest to hand, because the evidence that would let anyone speak with more precision is thinner than most people assume, and thinner in ways that are not evenly distributed.

Burnout is a real and recognised phenomenon. It is also, on close inspection, a category built mostly from research on men, applied to women's lives without much adjustment, and then quietly assumed to mean the same thing for everyone. This article sets out what is actually known, what is not, and why the shape of the gap matters as much as its size.

Burnout was named an occupational phenomenon, not an illness

In 2019 the World Health Organization added burnout to the eleventh revision of the International Classification of Diseases, but not as a medical condition. The entry is explicit about this: burnout is "a syndrome conceptualized as resulting from chronic workplace stress that has not been successfully managed," characterised by exhaustion, growing mental distance or cynicism about work, and reduced professional efficacy (World Health Organization, 2019). It sits in a chapter reserved for reasons people contact health services that are not themselves classified as illness, and the WHO was careful to say it should not be used to describe experiences outside the occupational context.

That distinction is easy to lose in everyday conversation, where burnout has come to describe everything from a bad quarter to a life in need of complete overhaul. It matters here because a syndrome defined around chronic workplace stress will inherit the blind spots of whatever workplace research produced it, and for most of the twentieth century that research population was disproportionately male, salaried and full-time. Women's occupational stress, layered as it usually is with unpaid labour outside paid hours, does not map cleanly onto a model built from that population.

The evidence on sex differences is smaller and messier than it looks

The most commonly cited answer to "do women burn out more than men" comes from a 2010 meta-analysis that pooled 409 effect sizes from 183 studies (Purvanova & Muros, 2010). Its finding is more careful than the headlines that followed it. Women showed slightly higher emotional exhaustion than men, a small effect. Men showed somewhat higher depersonalisation, the cynical-distance component, also a small effect. The authors explicitly warned against using their results to support the popular belief that women are more burnout-prone overall; the two sexes appear to experience different facets of the same syndrome, not different amounts of an identical one.

What the meta-analysis could not settle is why. It was built from studies designed to measure burnout, not from studies designed to trace the daily conditions that produce it, so the effect sizes describe an outcome without much of a mechanism attached. Later, smaller studies have tried to fill that gap with more texture: some report that women with burnout have more impaired awakening and lower job control, and describe unpaid work as a larger proportion of their week, while men with burnout more often describe overtime and narrowed social contact. These are single findings rather than a settled picture, and the Observatory regards them as leads worth watching rather than settled fact.

Great Britain's own occupational statistics show the practical scale of the problem without settling the sex-difference question either. Work-related stress, depression or anxiety accounted for 22.1 million lost working days in 2024/25, the single largest driver of work-related ill health recorded by the Health and Safety Executive, with those affected losing an average of nearly 23 days each (Health and Safety Executive, 2025). The headline figures are not broken down by sex in the same release, which is itself a small illustration of the gap this article is about: the number that would let researchers ask whether the burden falls unevenly is not the number that gets published first.

A second shift of unpaid care rarely appears in the numbers

Whatever burnout research measures inside the workplace, most of it stops at the office door, and that boundary falls differently across sexes. Analysis of English and Welsh data found that 57.7 per cent of unpaid carers were women and 42.3 per cent were men, and that among people who were also in paid employment, 13.3 per cent of economically active women were providing unpaid care compared with 9.5 per cent of economically active men, a gap that widened further in the fifty-to-sixty-four age band (Office for National Statistics, 2013). A separate, more recent systematic review of thirty-one peer-reviewed studies on the cognitive side of household work, the planning, anticipating and remembering rather than the physical doing, found that this "mental load" falls consistently more heavily on women, is shared more evenly only when both partners work full-time, and is linked in several of the reviewed studies to higher stress and anxiety (Reich-Stiebert, Froehlich & Voltmer, 2023).

The timing compounds the problem. The fifty-to-sixty-four age band, where the Office for National Statistics found the widest sex gap in unpaid care, overlaps closely with the years in which perimenopause and menopause typically occur, and with the working decades the Health and Safety Executive's overall figures describe as the most affected by work-related stress, depression or anxiety across the whole labour force. None of these datasets was built to talk to the other two, so a woman living through all three pressures at once rarely appears as a single, coherent case in anyone's evidence base. She appears instead as three partial entries in three separate systems, each measuring a slice of her week and none measuring the whole of it.

Neither of these bodies of evidence was designed as burnout research. That is the point. A woman's capacity in a given week is being drawn down by two systems, paid work and unpaid care, that are measured by two different disciplines using two different instruments, and almost nobody is adding them together. The Observatory's interest in longitudinal evidence comes directly from this problem: a single snapshot of workplace stress cannot see the second shift that started as soon as the first one ended.

Research on women's bodies has trailed research on men's for decades

The thinness of the evidence is not only about what happens outside the workplace. It runs through the biomedical literature that would, in principle, explain why exhaustion, sleep disruption and cognitive fog cluster the way they do in midlife women. A 2022 analysis of a century of biomedical publishing found that studies including both sexes remained around fifteen per cent of the literature throughout that period, that fewer than half of those studies went on to analyse their data by sex even when both were included, and that some fields, neuroscience and pharmacology among them, published five times as many male-only studies as female-only studies, or more, as recently as 2009 (Zucker, Prendergast & Beery, 2022). The same analysis found that women experience adverse drug reactions at markedly higher rates than men, a pattern that tracked closely with which drugs had female-biased pharmacokinetics, meaning some of the harm is plausibly a direct consequence of dosing regimes tested mostly on male physiology.

Funding follows the same pattern. A 2025 review by the National Academies of Sciences, Engineering, and Medicine found that women's health research received 8.8 per cent of all US National Institutes of Health grant spending between 2013 and 2023, a proportion that had fallen further, to 7.9 per cent, by the end of that window, despite women being roughly half the population the agency exists to serve (National Academies of Sciences, Engineering, and Medicine, 2025). Conditions that disproportionately affect women, the report noted, have not been funded in line with the burden they cause. None of this proves that burnout itself is under-researched in women specifically. It shows the surrounding terrain, the sleep science, the hormonal science, the pharmacology, that any account of women's capacity would need to draw on is systematically less developed than the equivalent terrain for men.

What the Observatory observes

None of this evidence lets anyone say precisely how women's burnout differs from men's, and the Observatory is wary of anyone who claims otherwise. What it does support is a narrower, more useful observation: the standard instruments were built on incomplete foundations, so a single questionnaire administered once is unlikely to capture what is actually happening in a woman's week.

This is why the Observatory's method leans toward pattern over snapshot. Members are invited to record simple, self-reported observations, sleep, energy, stress, everyday context, across weeks and months rather than at a single sitting, and to try small, optional changes and note how they felt, entirely at their own pace. Contribution to the aggregated, anonymised dataset behind this is a separate and optional choice, made with clear consent, and nothing here amounts to a claim of results. The point is narrower than that: a longer, more honest record is more likely to show where capacity is actually being spent than any single measurement can, precisely because the surrounding evidence base is too patchy to be trusted to fill in the rest. Readers who want the underlying reasoning in more detail may find what cognitive overload looks like in real life a useful companion to this piece.

The gap in the evidence is not a reason to stop looking. It is a reason to look differently, over more time, with the specific conditions of a woman's week left in view rather than assumed away. Observing your own pattern, honestly and without a predetermined story attached, is a small act, but it is one that the existing research base has rarely made room for.