How often should you get up from your desk? What the studies actually tested

Search for how often you should get up from your desk and you'll find confident numbers: every 30 minutes, every hour, every 20 minutes on the dot. Most are presented as if someone official settled the question. Nobody has. What we do have is a handful of studies that each picked an interval and tested it. That's more useful than a made-up rule, as long as you know what each one measured.

There's no official number

The World Health Organization's 2020 guidelines are the closest thing to an official answer, and they're careful here. They recommend that adults limit the time they spend sitting and replace some of it with activity of any intensity. But the evidence wasn't strong enough to set a threshold, so there's no guideline that says to stand up every so many minutes (Bull et al., 2020).

So when an article tells you the guideline is every 30 minutes, it's describing someone's reading of the research, not the research itself. Fair enough. Let's look at the research.

Every 20 minutes: the lab study

One of the most cited studies on breaking up sitting brought 19 overweight or obese adults, aged 45 to 65, into a lab and had them sit for five hours on three separate days (Dunstan et al., 2012). On one day they sat without a break. On the others they got up every 20 minutes for a two-minute walk, at either a light or a moderate pace.

After a standard test drink, blood sugar and insulin rose less on the walking days than on the day without breaks. Light walking worked nearly as well as moderate walking.

Keep the limits in view. It was 19 people on single days in a lab. The breaks were walks, not stretches. And the outcome was a blood marker over a few hours, not long-term health. It tells you that frequent short walks can change something measurable within a day. It doesn't tell you that every 20 minutes is the one right number.

Every 20 minutes again: computer workers and discomfort

A much older field study went into real workplaces instead (McLean et al., 2001). Computer workers did their usual keyboard work under different microbreak schedules and reported how uncomfortable they felt in the neck, low back, shoulders, and wrists and forearms.

The microbreaks reduced discomfort in every area studied, most of all when they came every 20 minutes, and there was no sign that productivity suffered. The study is small and dates from 2001, and the authors didn't claim it prevents injuries. It's still one of the few that compared intervals directly.

Every 30 minutes: cutting total sitting

A Cochrane review of workplace interventions for sitting less covered a much wider body of research (Shrestha et al., 2018). Most of it was about sit-stand desks, but one study compared breaks of one to two minutes every 30 minutes with two 15-minute breaks a day. The frequent short breaks cut sitting by about 40 minutes a day more. That's a single study, rated low quality, and the outcome was sitting time rather than health. But it points the same way as the others: short and often seems to fit a workday better than long and rare.

So what should you do?

Put it together and a pattern appears. The studies that tested a rhythm mostly tested something between every 20 and every 30 minutes, with breaks of one to a few minutes. None of them pitted every 20 minutes against every hour in desk workers over months, so treat the interval as a starting point, not a rule.

• If you can manage it, a minute or two of movement every 20 to 30 minutes is what most of these studies actually tested.
• If that's not realistic, fewer breaks still count. The WHO's point is that some activity is better than none.
• Attach breaks to things that already happen: the end of a call, a finished task, a water refill.
• Make them different from sitting. Stand up, walk, or move the parts of you that have been still.

DeskBreak defaults to three minutes for a practical reason: that's long enough to move your neck, shoulders, back, wrists, hips and legs, and short enough that you'll do it again later. How often you take one is up to you, and the studies above are a reasonable guide.