How Often to Get Up From Your Desk: What the Break Research Found

The advice to take breaks is universal and almost never quantified. How many, how long, and what happens to your work if you take them are answerable questions, and two field studies answered them fairly precisely.

The answer is smaller than most productivity advice suggests, and the part that will matter most to anybody negotiating with a manager is what happened to output.

The study

Galinsky and colleagues ran a field study with data entry operators, published in the American Journal of Industrial Medicine in 2007 as a follow-up to earlier work.

The design was simple. Workers spent four weeks on their conventional break schedule, which was two 15-minute breaks per day. Then they spent four weeks on a supplementary schedule: the same two 15-minute breaks, plus four additional five-minute breaks spread through the day. Fifty-one workers completed the symptom questionnaires.

Discomfort and eyestrain were both significantly lower with the supplementary breaks. The breaks also attenuated the accumulation of discomfort and eyestrain across a work session, meaning the usual climb through the afternoon was flattened rather than just shifted.

And the finding that matters institutionally: this happened without impairing productivity. Twenty extra minutes of break per day, and output held.

That is a converging result rather than a single one. The 2007 study was explicitly a follow-up to earlier field work reaching the same conclusion, which is a stronger position than one study on its own.

What the same study found about stretching

Here is the part usually left out, and it is worth being straight about.

Half the participants were assigned to a stretching group, asked to perform stretches during their breaks. Compliance was poor: workers reported stretching during only 25% of conventional breaks and 39% of supplementary breaks. No significant effects of stretching on discomfort or performance were observed.

The authors were clear that the low compliance prevented a valid assessment of whether stretching works. So the correct reading is not "stretching does nothing", it is "this study could not tell you, because people did not do it."

Which is itself informative. If a supervised workplace study cannot get people to stretch during more than 39% of their breaks, a plan built on stretching at your desk is a plan with a compliance problem before it starts. The break itself is the intervention that was demonstrated. The stretching was the part that did not happen.

The schedule the evidence supports

Schedule	Breaks per 8-hour day	Evidence	Effect on output
Conventional	Two 15-minute breaks	The comparison condition	Baseline
Supplementary (tested)	Two 15-min plus four 5-min	Field study, significantly lower discomfort and eyestrain	No impairment measured
Pomodoro (25 on, 5 off)	About 13 breaks	A productivity technique, not tested for musculoskeletal outcomes	Not measured in this literature
20-20-20 (eyes)	Every 20 minutes, look 20 ft away for 20 seconds	An eye-strain convention, widely repeated, weakly evidenced	Not measured

So: four extra five-minute breaks, on top of whatever you already take. Roughly one every ninety minutes across a working day.

That is a low bar, and it is deliberately a low bar, because it is the one that was measured and held up.

What to do with five minutes

The study did not test what people did with the breaks, only that they took them. Which means anything below this line is mechanism rather than evidence, and should be read as such.

The one thing worth doing is changing position rather than changing screen. A five-minute break spent scrolling a phone at your desk keeps your neck flexed, your shoulders rounded and your eyes at the same focal distance, which is most of the exposure you were breaking from.

Stand up. Walk to another room. Look at something more than twenty feet away. That covers the postural exposure, the static loading and the accommodative demand at once, and it requires no discipline beyond leaving the chair.

If you want to do something more structured with the time, five upper-back stretches take about three minutes, with the compliance caveat above firmly in mind.

Why breaks matter more than the fix you were about to buy

This is the cheapest intervention in the entire category and the one with the most direct field evidence behind it. It costs nothing. It needs no product. It was measured in a real workplace against real output.

Compare that to the alternatives. The trial evidence for a posture corrector shows reduced trapezius activity while worn, with no significant change in the posture angle. The best standing desk trial got occupational sitting down by 83 minutes a day after a year of effort. Both are real findings and both took a purchase.

The break schedule took twenty minutes a day and no money, and it moved discomfort and eyestrain significantly.

None of these are alternatives to each other. A standing desk changes how long you hold one position, breaks change how often you leave it, and the geometry of the screen changes how bad that position is in the first place. But if you are doing exactly one thing, the free one has the most direct evidence.

Why the breaks work, as far as anybody can tell

The study measured outcomes rather than mechanisms, so what follows is inference and should be read as such. Three things plausibly change when you leave the chair for five minutes.

The static load stops. Sitting at a screen is not a rest position. The muscles holding your head over your spine and your shoulder blades against your ribcage are working the whole time, at a low level, without a break. Low-level sustained contraction is a recognised problem in the occupational literature precisely because it never lets the tissue recover, and standing up interrupts it.

Your eyes change focal distance. Eyestrain fell in the study alongside discomfort, and the most obvious candidate is that a break gets your eyes off a fixed focal plane at fixed brightness. This is also why a break spent on a phone at your desk does considerably less than a break spent walking to another room.

The position changes at all. The strongest single finding across this whole subject is not about any particular posture, it is about duration in one posture. The best sit-stand trial got its result by reducing time seated rather than by improving how people sat, and the break literature reaches the same place from a different direction.

That third point is worth sitting with, because it reframes the whole category. Most posture advice is about finding the correct position. The evidence keeps pointing at not staying in any position for very long.

What this does not tell you

Two limits, stated plainly.

The study population was data entry operators, whose work is more repetitive and more continuously seated than most desk jobs. Whether the same schedule produces the same effect for somebody who already gets up for meetings is not something the study can answer, and somebody who is already interrupted six times a day may be getting the intervention without knowing it.

And the outcomes were self-reported discomfort and eyestrain over a four-week block. That is a real measurement and it is not a measurement of injury, of long-term musculoskeletal outcomes, or of anything happening to the tissue. The claim supported here is that people felt better and worked the same amount, which is a good claim and a modest one.

Making it happen

The failure mode is not disagreement, it is that nobody remembers. Three approaches that work, in rough order of how well they survive a busy week:

• Attach the break to something that already happens. End of a meeting, before a call, when a build runs. A break with an existing trigger needs no willpower.
• Put it in the calendar as four five-minute blocks. Crude, and it survives contact with a shared calendar better than an intention does.
• A timer. Works for about a fortnight for most people, then gets dismissed automatically. Useful for establishing the pattern, unreliable for maintaining it.

Whatever the mechanism, the target is four, five minutes each, out of the chair. That is the whole prescription.

Questions people actually ask

How often should you take a break from your desk?

The field study that measured it added four five-minute breaks on top of two conventional 15-minute ones, which is roughly one extra break every ninety minutes. Discomfort and eyestrain fell significantly and productivity was not impaired.

Do short breaks reduce productivity?

Not in the study that measured it. Workers took twenty extra minutes of break per day and output held, which is the finding worth quoting to anyone who needs convincing. The result replicated earlier field work reaching the same conclusion.

Should I stretch during breaks?

The same study assigned half its participants to stretch and found no significant effect, but compliance was so low, 25% of conventional breaks and 39% of supplementary ones, that the authors said it prevented a valid assessment. The break itself is the part that was demonstrated.

Is the Pomodoro technique good for posture?

It produces frequent breaks, which is the mechanism that was tested, so it is likely to help by accident. It was designed as a focus technique and has not been evaluated for musculoskeletal outcomes, so treat any posture benefit as plausible rather than measured.

What should I actually do in a five-minute break?

Leave the chair and look at something far away. That changes your posture, unloads the static holding and changes your eyes' focal distance at the same time. A break spent scrolling a phone at your desk preserves nearly all of the exposure you were breaking from.

Sources

• Galinsky T et al., "Supplementary breaks and stretching exercises for data entry operators: A follow-up field study", American Journal of Industrial Medicine, 2007. PubMed 17514726
• Galinsky T et al., "A field study of supplementary rest breaks for data-entry operators", Ergonomics, 2000. PubMed 10877480
• "Work-break schedules for preventing musculoskeletal symptoms and disorders in healthy workers", Cochrane Database of Systematic Reviews, 2019. PubMed 31334564
• Edwardson et al., "Effectiveness of the Stand More AT (SMArT) Work intervention: cluster randomised controlled trial", BMJ, 2018. PMC6174726