Long working hours and heart health: what the research can tell us

How many hours are you actually working?

Start with a representative week, including the work that never appears on a rota: evening messages, preparation, invoices, weekend catch-up and a second job. Keep commuting and caring time in separate columns. They affect the time available for rest, but separating them avoids quietly changing what your work-hours total means.

Here is a simple example. A contracted 40-hour week plus 45 minutes of work on five evenings adds 3 hours and 45 minutes. Another two hours on Sunday makes 45 hours and 45 minutes in total. The extra work may arrive in small pieces, but the weekly arithmetic is straightforward. Writing it down can make a specific discussion possible.

Then identify which part is changeable. Is there a recurring task that could stop? Are two people producing the same report? Does every evening message require an answer that night? Is a second job essential to meet current costs? The last question may reveal a financial constraint that advice to set better boundaries simply ignores.

Take the most concrete problem to someone who can act on it: a manager, occupational-health team or trade-union representative where available. HSE encourages these conversations when work is causing stress. HSE guidance provides a starting point for both workers and managers.

If you work for yourself, the equivalent conversation may be about capacity and pricing, deadlines or which work you accept. Choose one change whose effect you can actually see in the next week's diary. An earlier finish on paper is not much use if the same task reappears after dinner.

How could working time affect your heart?

The WHO describes two broad pathways. One involves repeated stress responses, including changes in stress hormones and nervous-system activity that can affect blood pressure and cardiovascular function. The other involves behaviour and recovery: less opportunity for sleep and activity, alongside possible changes in smoking, alcohol use and diet.

These are plausible routes through which work can influence health, rather than proof that every person working long hours follows the same sequence. Someone may enjoy the job yet have little time left to sleep; someone else may work fewer hours but face intense demands and little control. Hours are an important part of the picture, but they are not a complete description of working conditions.

The UK's Health and Safety Executive identifies six areas relevant to work-related stress: demands, control, support, relationships, role and organisational change. It says employers should assess and act on stress risks. That gives a workload discussion more substance than simply asking whether someone feels resilient enough to cope.

For example, a person may be able to finish a normal day's work until an additional reporting task arrives. If that task has no allocated time, it moves into the evening. A wellbeing session does not resolve the scheduling conflict. The practical question is what will be removed, delayed, shared or simplified so that the work fits into the available time.

What did the major research reviews find?

The 2020 heart-disease review pooled 22 studies involving 339,680 people for the comparison of new ischaemic heart disease at 55 or more versus 35–40 weekly hours. The estimated relative risk was 1.13, with a 95% confidence interval of 1.02–1.26: an estimated 13% increase, with uncertainty spanning roughly 2–26%. For death from ischaemic heart disease, the separate estimate was 17% higher, from 16 studies involving 726,803 people. The authors rated both bodies of evidence moderate quality. Li and colleagues

Ischaemic heart disease concerns an inadequate blood supply to the heart muscle. Developing the disease and dying from it are different outcomes. Keeping their estimates separate prevents a common problem in headlines: a percentage that loses its meaning when the event being counted is left out.

The stroke review found a relative risk of 1.35 for new strokes in the same hours comparison, with a 95% confidence interval of 1.13–1.61. That analysis included seven studies and 162,644 people and was rated moderate quality. Evidence on death from stroke was less certain and did not show a clear increase. The 35% figure therefore describes stroke occurrence, not stroke deaths.

The stroke association was larger than the heart-disease association in those analyses, but the estimates came from different sets of studies. They cannot tell an individual worker which event is more likely. Absolute risk also depends on the person's starting risk and the period being considered.

The reviews mainly draw on observational evidence: researchers follow people with different working patterns instead of randomly assigning years of overtime. Combining studies can make an estimate more informative, but it does not remove differences in jobs, income, smoking, health or other circumstances. Adjusting for measured factors helps; factors that were missed or measured poorly can still matter. Heart-disease review methods

How large is the risk in absolute terms?

A later US study offers useful context. It followed 4,051 employed adults without a previous heart attack or stroke at the 1995–96 starting assessment, with deaths tracked through 2021. Observed cardiovascular-death rates were 2.74 per 1,000 person-years among people reporting at least 55 weekly hours, compared with 1.86 among those reporting 35–40 hours. MIDUS cohort study

A person-year means one person followed for one year; ten people followed for a year contribute ten person-years. The figures above are descriptive, unadjusted rates. The study's separate adjusted hazard ratio was 1.50, with a confidence interval of 1.03–2.17. That describes a higher relative rate during follow-up, not a 50% chance of cardiovascular death. Working hours were recorded once, so changes in hours over later decades were not captured. Original study

The distinction between those two results matters. You cannot subtract the observed rates and call the difference the number of deaths caused by overtime after accounting for every other factor. Nor can you multiply one rate by ten to produce your own ten-year forecast. The numbers are useful because they show both the scale of events and the uncertainty behind a striking relative percentage.

When reading the next headline, look for three things together: what happened, what it was compared with, and over what period. “Risk rose by half” is incomplete until you know those details. An estimate about cardiovascular death over long follow-up answers a different question from one about a short-term change in blood pressure or wellbeing.

Does a four-day week solve the problem?

A 2025 study in Nature Human Behaviour examined 2,896 employees across 141 organisations in six countries during a six-month reduced-hours trial with pay preserved. The organisations reorganised work before the trial. Reported wellbeing improvements included burnout, job satisfaction, and mental and physical health; corresponding improvements were not seen in 12 control companies. These were wellbeing outcomes, not measurements of heart attacks or stroke deaths.

The distinction between reducing and compressing hours is crucial. Four ten-hour days still make 40 hours. Four eight-hour days make 32. A policy can change the number of commuting days without reducing the total work, while another can reduce hours through changes to how work is done. The label four-day week does not tell you which arrangement is being offered.

The study is a reason to examine reduced-hours working seriously, but its six-month wellbeing findings do not supply a cardiovascular-risk reduction percentage. For a proposed change at your workplace, ask what happens to total hours, workload, pay and out-of-hours contact. A nominal day off that fills with unfinished work is a different intervention from a real reduction.

Make sure recovered time stays available

Use the time you recover deliberately. If rotating shifts are the problem, the shift-work sleep guide deals with sleep opportunities around a changing schedule. If a long seated day leaves little room for movement, the exercise-snacks guide explains the evidence for short activity bouts. Neither gives a formula for cancelling out a particular number of overtime hours.

Keep the workload change and the personal routine visible as separate actions. You might agree to remove a late meeting and use the recovered time for a regular meal or walk. Recording both makes it easier to notice when the meeting returns, instead of concluding that you have failed to maintain a healthy routine.

Acute symptoms need a different response. Call 999 for chest pain that feels tight or squeezing, pain spreading to the arms, neck or jaw, or severe breathing difficulty, as set out in NHS heart-attack guidance. For sudden facial weakness, arm weakness or speech problems, use FAST and call 999. NHS stroke guidance says to seek emergency help even if stroke signs within the last 24 hours have stopped. Do not drive yourself to A&E.

The research supports treating long working hours as a health issue that deserves action. The first useful result is often modest and specific: an honest total, an agreed reduction in a recurring demand, and recovery time that remains available when the next busy week arrives.