Quick answer: WHO and ILO joint estimates found that working 55 or more hours a week is associated with an estimated 35% higher risk of stroke and 17% higher risk of dying from ischaemic heart disease, compared with working 35–40 hours. That is a cardiovascular finding, not a burnout diagnosis — but it is the clearest public-health marker available for where sustained load stops being benign.

Working hours are the least glamorous variable in every wellbeing conversation and usually the most decisive one. They are also the one people most consistently misreport to themselves, because the count feels like a judgement rather than a measurement.

What the WHO and ILO actually published

In 2021 the World Health Organization and the International Labour Organization released joint estimates of the burden of disease attributable to long working hours. The headline exposure was 55 hours or more per week, compared against a reference of 35–40 hours. The associated increases: an estimated 35% higher risk of stroke and 17% higher risk of dying from ischaemic heart disease.

Three things worth holding onto:

Why hours do damage: two mechanisms

Direct load

Sustained demand keeps the sympathetic nervous system engaged and the parasympathetic brakes underused. A systematic review in PLOS ONE found job stress and burnout associated with reduced heart rate variability, specifically lower parasympathetic activity. That is the physiological signature of a system that does not get to stand down.

Displacement

The subtler mechanism, and the one you can actually observe. A 55-hour week does not add hours to the day; it takes them from somewhere. Almost always: sleep, movement, and unstructured recovery.

The consensus recommendation for adults is seven or more hours of sleep per night on a regular basis. A long week plus a commute plus any life at all makes that arithmetically hard, and the deficit is where much of the harm gets delivered.

Count your real week

Most people undercount by five to fifteen hours, because they count time at a desk and nothing else. A truer count includes:

Usually counted Usually not counted — but should be
Hours at the desk or on siteEmail and messages on the commute
Scheduled meetingsEvening triage: "just twenty minutes" after dinner
Explicit overtimeSunday planning or catch-up
Client workOn-call time you cannot disengage from
Work travel outside normal hours

The test is not "was I paid for it" but "was my attention occupied by work". On-call counts even when nothing happens, because being unable to switch off is the load.

What to do at each level

Real weekly hours Reasonable posture
Under 45Hours are probably not your problem. Look at control, support and sleep instead.
45–55Sustainable for a season, not a year. Protect sleep aggressively and watch whether days off still restore you.
55+ occasionallyA crunch. Give it an end date, in writing, and a recovery period after it.
55+ as the normThis is the WHO/ILO exposure category. Treat reducing it as a health decision, not a career preference.

Hours are not the whole story

Two people at 55 hours can be in completely different situations. A systematic review with meta-analysis in BMC Public Health found low workplace support and low job control associated with burnout symptoms, with job support protective for emotional exhaustion. Long hours you chose, in work you control, with a manager who absorbs shocks, are not the same as long hours imposed on you with neither.

That is why a serious picture needs hours and autonomy and support and what your sleep is doing. Any one of them alone will mislead you.

The practical version

Log the hours you actually work

Pulse takes your work hours as one of fourteen signals and shows how they move with your sleep, HRV and resting heart rate. Free, on-device.