Long working hours and stroke: working 55 hours a week was tied to a one-third higher risk

In May 2021 the World Health Organization and the International Labour Organization put a number on overwork. Long working hours, they announced, led to 745,000 deaths from stroke and ischemic heart disease in 2016. “Working 55 hours or more per week is a serious health hazard,” said Maria Neira, who directs the WHO’s Department of Environment, Climate Change and Health.

That figure rests on a body of research that is narrower and more cautious than the headline. A central study behind it appeared in The Lancet in 2015. Its lead author has since argued in print that the WHO went further than the evidence allows.

What the 2015 analysis of long working hours measured

Mika Kivimäki of University College London and colleagues set out to settle a question on which the published studies were few and imprecise. They searched the literature, and they also obtained unpublished data for 20 cohort studies from a research consortium and from open data archives. Unpublished data matter because studies that find nothing are less likely to reach print, and an analysis built only on published results can overstate an effect.

The final pool comprised 25 studies from 24 cohorts in Europe, the USA, and Australia. For heart disease it covered 603,838 men and women who were free from coronary heart disease at baseline, and for stroke 528,908 who had never had one. Participants were free of the disease in question at the start and were then followed, for about eight and a half years on average in the heart analysis and a little over seven in the stroke analysis. In that time 4,768 coronary events and 1,722 strokes were recorded.

The researchers compared people working 55 hours or more a week with those working a standard 35 to 40.

What the analysis found about long working hours and stroke

After adjustment for age, sex and socioeconomic status, people working 55 hours or more had a risk of stroke about one-third higher than those on standard hours. Their risk of coronary heart disease was 13% higher. The authors’ own summary gave the two results different weight: “Employees who work long hours have a higher risk of stroke than those working standard hours; the association with coronary heart disease is weaker.”

Two features made the stroke finding more persuasive than a single comparison would be. The first was a gradient across the bands of hours.

Weekly working hours	Stroke risk compared with 35 to 40 hours
41 to 48	About 10% higher, not distinguishable from no difference
49 to 54	About 27% higher
55 or more	About 33% higher

The second was that the result survived the obvious objections. The excess risk of stroke remained unchanged in analyses that addressed reverse causation, multivariable adjustments for other risk factors, and different methods of stroke ascertainment. Reverse causation here means the possibility that early, undiagnosed illness changes how much people work, which would distort the comparison.

A sense of scale helps. Strokes were uncommon in these cohorts of working-age adults: 1,722 among 528,908 people over roughly seven years comes to about three in every thousand. A one-third rise on a base that small is a modest change in any one person’s chances, though across a large workforce it adds up.

Why long working hours might lead to stroke

Two routes are usually proposed, and the WHO analysts set them out. The first is through biological responses to psychosocial stress, in which the release of excessive stress hormones may disturb the cardiovascular system. The second pathway is through behavioral responses to stress that are established cardiovascular risk factors, including tobacco use, harmful alcohol use, unhealthy diet, physical inactivity, and impaired sleep.

Kivimäki’s group tested one specific possibility in 2017. Atrial fibrillation, an irregular heartbeat, is the most common cardiac arrhythmia and a risk factor for stroke. Among 85,494 working men and women followed for ten years, individuals working long hours had a 1.4-fold increased risk of atrial fibrillation compared with those working standard hours. That offers a plausible bridge from long hours to stroke.

It does not explain how long hours would produce the arrhythmia. Smoking, obesity, inactivity and heavy drinking were only slightly more common among those working long hours, by less than three percentage points, and blood pressure and cholesterol did not differ. The authors concluded that classic risk factors are unlikely to account for the link, and stated that the mechanisms underlying the association between long working hours and atrial fibrillation are not known.

How the 745,000 figure for long working hours was built

The WHO and ILO commissioned their own reviews of the evidence, one for each disease. The stroke review, led by Alexis Descatha, covered 22 studies with a total of 839,680 participants in eight countries. It reached almost the same estimate as the Lancet paper, a risk of stroke about 35% higher at 55 hours or more, and rated the evidence as moderate quality. For 49 to 54 hours it found an increase of about 13%. On the question of dying from a stroke, as distinct from having one, the reviewers wrote that they were very uncertain about the effect.

The heart disease review, led by Jian Li, covered 37 studies comprising a total of 768,751 participants in 13 countries. It estimated that 55 hours or more raised the risk of developing ischemic heart disease, another term for coronary heart disease, by about 13% and of dying from it by about 17%, and it judged the data on working 55 hours or more to be sufficient evidence of harmfulness for both.

A third paper, led by Frank Pega of the WHO, turned those risks into deaths. The method multiplies three things: the share of each country’s population working 55 hours or more, taken from national surveys; the extra risk from the two reviews; and the total number of deaths from each disease. On that basis the authors estimated that 488 million people, 8.9% of the global population, were working such hours in 2016, and that 745,194 deaths from ischemic heart disease and stroke combined were attributable to this exposure. That represents 3.7% of all deaths from ischemic heart disease and 6.9% of all deaths from stroke.

According to the agencies’ announcement, 72% of the deaths occurred among men, and the burden fell most heavily on people living in the Western Pacific and South-East Asia regions.

Such a calculation depends on its assumptions, and the paper states them. The authors write: “Evidence comes from observational studies, for which residual confounding cannot be ruled out, and we cannot be certain that a causal association exists.” The method also applies risks measured in a limited set of countries to the whole world. The heart disease review drew on studies from the Americas, Europe and the Western Pacific, and none from South-East Asia, the region where long hours were found to be most common.

Why the heart disease part of the claim is disputed

The sharpest criticism came from Kivimäki and three colleagues, in a commentary titled to leave no doubt: the conclusions, they wrote, are not supported by the evidence.

Their argument concerns heart disease and not stroke. Looking inside the WHO/ILO review, they found that the association between long working hours and heart disease appeared stronger in lower quality cohort studies with a high risk of bias than in the better studies, where the estimate could not be told apart from no effect. They also found that the association depended on the kind of job. There was no increase in ischaemic heart disease for those working long hours in high socioeconomic status occupations. The WHO/ILO reviewers had seen something similar, reporting that risks were higher among persons with lower SES, meaning lower socioeconomic status.

The commentary ends on a restrained note. “The conclusions should be restricted to low socioeconomic status occupations only and more research is still needed to confirm or refute harmfulness and determine clinical relevance.” The WHO paper acknowledges the disagreement, noting that other researchers have disputed the rating of the evidence.

What else long working hours have been linked to

A broader study from the same European consortium, led by Jenni Ervasti of the Finnish Institute of Occupational Health, checked long hours against 50 health outcomes at once in more than 100,000 workers from Finland, Sweden, Denmark and the United Kingdom. Looking at many outcomes together guards against reporting only the striking ones.

It found a mixed picture. Working long hours was associated with an increased risk of cardiovascular death and of infections serious enough to need hospital treatment before age 65, and, more weakly, with diabetes, injuries and musculoskeletal disorders. Against that, the authors reported plainly that working long hours was not associated with death from all causes.

What the evidence on long working hours cannot show

Every study in this field is observational. Nobody has been randomly assigned to a 55-hour week, and people who work such hours differ from those who do not in income, occupation, family circumstances and much else. Adjustment narrows those differences and cannot remove them.

A second weakness lies in measurement. In both WHO/ILO reviews the exposure was measured using self-reports in all studies, and in the European consortium’s cohorts working hours were assessed once, at the start, although hours change over a career. A person counted as working long hours may have done so for one year or for thirty.

The dependence on job status may be the most important loose end. Fifty-five hours in a high-status occupation and fifty-five hours in a low-status one are counted as the same exposure, and the evidence suggests they do not carry the same risk.

Kivimäki and colleagues closed their 2015 paper with a practical suggestion that stopped short of prescribing shorter hours: that more attention should be paid to the management of vascular risk factors in individuals who work long hours.

People who work 55 hours or more a week have had more strokes than people on standard hours in every large analysis so far, while whether the hours themselves are the cause, and how many deaths follow from them, is still argued over.