For a period, I worked during the day from Monday to Friday, with an occasional night shift inserted between those shifts. In practice, there were times when I had no conditions for proper sleep for more than two days. While travelling, I experienced memory lapses and perceptual disturbances, including hallucinations. I also associate my chronic headaches with night work and the constant disruption of my circadian rhythm.

I am not presenting this as a medical diagnosis or a public accusation against a particular company. The temporal connection I observe is not yet proof of causation. What interests me is a mechanism that is easily lost in a table of working hours: night work is not simply work that starts at a different time. Its cost also depends on what came before it, what follows it, and whether there is a real opportunity for sleep and normal life between shifts.

Three different problems hidden under one label

Night work, shift work and sleep deprivation do not mean the same thing. Night work describes when duties are performed. Shift work describes a system of moving between different working hours. Sleep deprivation is a state in which a person does not get the sleep needed to recover. These phenomena can occur together, but their effects need not be identical in every situation.

Humans do not function during the day and sleep at night by accident. Almost every tissue and organ has its own circadian rhythm, and light and darkness are among its strongest regulators (NIGMS, “Circadian Rhythms”). People differ in chronotype and tolerance of night work, but the body cannot be reduced to a clock that can be moved forward or back by several hours at no cost.

NIOSH links nonstandard and long working hours to insufficient sleep, fatigue, poorer cognitive functioning and an increased risk of errors, injuries and accidents (CDC/NIOSH, “Fatigue and Work”). A review of studies on severe sleep deprivation describes perceptual disturbances, including hallucinations, that increase with time spent awake (Waters et al., 2018). A meta-analysis of seven observational studies also found an association between shift and night work and more frequent headaches (Wang et al., 2024). None of these sources diagnoses my case. They do show, however, that such symptoms cannot automatically be dismissed as a private inconvenience or an unwillingness to work.

A schedule can balance the table while taking sleep away

A schedule lets us count the beginning and end of every shift. It is harder to see the commute, eating, winding down, trying to fall asleep during the day and getting ready for work again. A formal break between two entries in a table is not automatically sleeping time.

Nor is it automatically free time. When I sleep until late afternoon after a night shift, I have to arrange my entire life around work. If I get up at 16:00, I can no longer deal with a matter at an office that closes at 16:00. Ordinary hours for relationships, responsibilities and recovery disappear in much the same way. The table shows a day without a shift, but in reality part of that day is the biological repayment of the previous night's cost.

That is precisely why a single night shift squeezed between a series of day shifts can be different from a predictable period of night work. A person cannot switch circadian rhythm like a setting in an app. When a night shift is followed by a quick return to daytime work, some of the nominal rest is consumed by the transition itself.

My proposal was simple: two workers could alternate whole weeks of night work instead of switching unpredictably between day and night. The research does not, however, allow me to honestly claim that a week-long block is optimal. NIOSH advises against quick returns between shifts and notes that forward rotation can be easier, but also identifies weekly rotation as particularly difficult. It remains uncertain whether fast forward rotation or slow rotation lasting at least two weeks is better (CDC/NIOSH, “Work Organization Strategies”).

A 2023 Cochrane review reached a similarly cautious conclusion: much of the evidence concerning rotation direction and speed, shift length and time between shifts is of low or very low certainty. It does not offer a single recipe for every profession and person (Hulsegge et al., 2023). This is an important correction. The problem exists, but an intuitive solution still needs to be tested in the specific working environment.

Long-term costs require honest language

I have heard the claim that night work shortens life by several years. I found no basis for repeating that as a general truth. A cohort study using data from 192,764 UK Biobank participants associated regular night work with accelerated biological ageing. For regular night workers aged 45, estimated life expectancy was on average 0.94 years lower than for day workers (Cai et al., 2025). That is a statistical association in the population studied, not proof that every night shift takes a specific number of days away from a particular person.

Vitamin D requires similar caution. The body can produce it in the skin under UVB radiation, so moving work into the night and sleep into the day can reduce opportunities to spend time in sunlight. A meta-analysis of 13 cross-sectional studies found lower average vitamin D levels among shift workers, but results varied considerably, and diet, body weight, season and other factors could also play a role (Coppeta et al., 2022). That is a reason to include health in preventive measures around night work, not to diagnose a deficiency oneself or attribute every symptom to it.

Careful language does not weaken the argument. Quite the opposite: we do not need to claim that all night work will shorten life by several years to recognise that repeated disruption of sleep, circadian rhythm and life outside work is a real cost that an organisation must manage.

Rest is part of the safety system

Not all night work can be eliminated. Hospitals, infrastructure, transport and many services must operate while most people sleep. Different people may also tolerate the same hours differently. Responsible organisation therefore does not mean declaring every night shift bad. It means designing schedules so that necessary work does not deprive people of the ability to act safely.

Switzerland's SECO treats the balance between work and rest as part of protecting health and preventing fatigue-related accidents. The general rule provides for at least 11 consecutive hours of daily rest. In systems with three or more shifts, rotation should generally move from morning to afternoon to night. Employees should be involved in planning working time, and new hours should be communicated as early as possible, usually two weeks before they take effect (SECO, “Arbeits- und Ruhezeiten”).

These are minimum legal safeguards, not a guarantee of recovery in every situation. Eleven hours between shifts does not mean eleven hours of sleep. A schedule should therefore be assessed not only for compliance with a limit, but also for commuting time, predictability, rotation direction, successive shifts and observed effects on the worker.

The Swiss framework shows something else: with regular night work, health is not solely the employee's private problem. People working at least 25 night shifts a year are entitled to a medical examination and consultation, and under certain conditions the examination is mandatory. SECO also points to the importance of recovery time and cumulative strain (SECO, “Medizinische Untersuchung und Beratung bei Nachtarbeit”). I am not determining who is entitled to a particular benefit in an individual case. I am illustrating a principle: if an organisation relies on work that runs against most people's natural rhythm, it must take part in managing its risks.

A person is not a hammer

A work schedule is part of a production system, but a person is not a hammer put back on a shelf after a shift. People do not instantly return to their initial state, and they do not exist only during the hours paid for by an employer. A poorly arranged transition between day and night can transfer an organisational cost into a worker's sleep, headaches, memory, attention, perception and safety. That same person then still has to get home, drive, run errands, build relationships and live.

An employee can contribute value over the long term only if work does not destroy the conditions needed to do it. Wearing down the body does not become a sensible model simply because it has been put into a schedule and paid for. Money can compensate for effort and time, but it cannot undo past sleep deprivation or guarantee the recovery of health. Lost time cannot be given back later either.

I do not know one ideal schedule for everyone. Research does not identify one either. We can, however, assess a system through concrete questions: does it limit quick returns, leave a real window for sleep, make shifts predictable, use a biologically sensible rotation, allow the employee to report effects, and actually respond to those reports?

Night work is sometimes necessary. Necessity does not make it neutral for health, and lack of sleep should not be treated as proof of availability. If a schedule works only because it disregards recovery and a person's life outside work, it is not the person who has failed the system. The system has been calculated badly—and is using the worker up instead of organising their work.

Bibliography

✦Building systems for greater autonomy.