How much exercise offsets sitting all day? Less than you have been told

That is the answer. Here is where it comes from and what it is worth.

Two studies, two numbers, one important difference

The famous number is from a harmonised meta-analysis published in The Lancet in 2016 by Ekelund and colleagues, pooling 16 studies and more than a million people. It found that the elevated mortality risk associated with eight hours of sitting a day was eliminated in people doing roughly 60 to 75 minutes of moderate-intensity activity daily. Brisk walking counts as moderate. That is the figure every headline used, and it is a discouraging one, because it is three to five times what most guidelines ask for.

The less famous number is better measured. In 2019, Ekelund and colleagues published a second harmonised meta-analysis, this time in the BMJ, restricted to studies that measured activity with accelerometers rather than questionnaires. Eight studies, 36,383 adults, mean age around 62. The largest reductions in mortality risk appeared at about 24 minutes a day of moderate-to-vigorous activity, or about 375 minutes a day of light activity.

Same lead author. Same question. A number less than a third the size.

The difference is not a contradiction, it is a measurement problem. People are bad at reporting their own activity, and the errors are not random. When you measure with a device instead, the associations get sharper and the apparent dose needed gets smaller. The 2019 paper notes its own risk reductions were about twice as large as those from self-reported studies.

If you want one number to act on, take the device-measured one. It is the more careful study, and it happens to be the more encouraging.

Where the sitting threshold actually sits

The two studies also disagree about how much sitting is too much, for the same reason.

• Patterson and colleagues, 2018, writing in the European Journal of Epidemiology, harmonised 34 cohorts totalling 1.33 million people, all using self-reported sitting. All-cause mortality risk began climbing above eight hours a day. Cardiovascular mortality began climbing above six.

• The 2019 accelerometer analysis put the inflection higher, above about 9.5 hours a day, and the gradient beyond it was steep: hazard ratios of 1.48 at ten hours and 2.92 at twelve.

Take the pair together and the practical reading is that somewhere between eight and ten hours is where this stops being theoretical, and that the far end of the range is where it gets serious. Most remote desk workers are inside that band. One Stanford Center on Longevity analysis put full-time remote workers at 9.2 sitting hours a day, with non-remote workers at 7.3, though it was self-reported and pandemic-era.

Why “sitting is the new smoking” is wrong, and why it matters

We refuse this framing, and not out of pedantry.

Vallance and colleagues dismantled the comparison in the American Journal of Public Health in 2018. Sitting eight or more hours a day is associated with a premature death risk somewhere around 10 to 20% higher. Smoking is associated with roughly 180% higher. Those are not adjacent quantities. One is a modest elevation in a large population; the other is the single most destructive consumer product ever sold.

Here is the part that matters more than the arithmetic. The slogan is not merely inaccurate, it is counterproductive, in two directions at once.

It produces fatalism. If your chair is a cigarette, you are already doomed by lunchtime and there is no obvious dose that saves you. That is not a message anybody acts on.

It produces the wrong intervention. The smoking analogy implies elimination. You cannot stop sitting, so people buy a standing desk, treat that as the fix, and go home. But the evidence is about accumulating movement, not about eliminating a substance. The right response to a modest, dose-related risk is a modest, repeated, cheap behaviour, which is a far easier thing to sell yourself on than quitting.

The honest version is less dramatic and more useful: this is a small risk you can reduce inexpensively, and the reduction appears at doses most people could reach without changing their life.

What “offset” is really doing in that sentence

The word implies a ledger. Sit for a debit, exercise for a credit, balance to zero. That is not what these studies found, and it is worth being precise about, because this article’s whole value is precision.

Three caveats, none of which cancels the finding.

• These are associations. Cohort studies observe people; they do not assign them to sit. The people who managed an hour of daily activity differed from those who did not in ways no statistical adjustment fully removes.

• “Eliminated” means the statistical association disappeared in that dataset, at that dose, for that outcome. It does not mean a physiological debt was repaid.

• Different outcomes behave differently. Mortality is one thing. Post-meal glucose is another, and there the evidence points at breaking up the sitting rather than at total daily exercise volume. In a small Columbia trial, a five-minute walk taken every half hour trimmed the post-meal glucose rise by 58%, and no other interval tested moved glucose at all.

That last point is the practical one. “How much exercise offsets sitting” and “how should I break up sitting” are two different questions with two different literatures, and you can act on both. We took the second one apart in how often should you get up from your desk.

So what should you actually aim at

Start with the official target, because it is the least contested thing on this page. The World Health Organization asks adults for 150 to 300 weekly minutes of moderate activity, or 75 to 150 if the effort is vigorous, plus strengthening work on two days or more. WHO also tells you to sit less while declining, deliberately, to specify how many hours or how often to break. Anybody who tells you WHO recommends standing up every thirty minutes has invented that.

Set against the device-measured finding, the arithmetic is friendlier than the headlines suggest. About 24 minutes a day of moderate activity is roughly 170 minutes a week, which sits inside the standard guideline rather than beyond it. On that reading, hitting the ordinary target is already close to the dose associated with the biggest risk reduction in people who sit a lot.

And the guidelines removed the barrier that used to make this impossible for desk workers. For decades activity was only credited when it arrived in unbroken ten-minute chunks. That requirement was dropped in the 2018 revision of the US guidance. Short efforts count now, and they count officially.

Twenty-four minutes, assembled from a working day

If you cannot find a continuous half hour, do not find one.

1. Six five-minute breaks across a working day is thirty minutes, and hourly is the pattern people actually sustain. In 2026, among more than eleven thousand desk workers, an hourly rhythm produced nearly the benefit of a half-hourly one and proved far more durable.

2. Make some of it brisk. Moderate means your breathing changes. Air squats, step-ups on a stair and marching all qualify; a slow wander to the kettle does not.

3. Put strength in two of them. Desk-edge push-ups and a row against a doorframe cover the muscle-strengthening half of the guideline, and in 27 randomised office-worker trials strengthening outperformed stretching for neck and shoulder discomfort.

4. Count the walk to the shops. Incidental activity is activity. The device-measured studies did not care whether it was exercise or errands.

5. Do not try to bank it at the weekend. These findings are about daily accumulation, and nothing here supports a Saturday repayment plan.

If assembling that yourself sounds like admin, that is what the ready-made plans are for. Full body strength covers the two strengthening days. Energy boost cardio is the one that raises your breathing in under five minutes.

What we are not claiming

Not that exercise breaks make you sharper. Studies that interrupted sitting and then measured thinking found nothing, and the pooled short-term cognitive return on exercise across 113 studies is g=0.13, resting on a literature where null findings were less likely to be published. The detail is in do exercise breaks improve concentration.

Not that short efforts fix your blood pressure or your body fat either. An 11-trial review in 2025 found brief repeated efforts improved cardiorespiratory fitness at moderate certainty and changed nothing in body composition, blood pressure or blood lipids.

What is defensible: movement reliably raises energy and mood, small doses build fitness, and taking breaks does not cost you output. That is the whole promise.

PomoFit is a browser Pomodoro timer that puts one guided bodyweight movement into every break, which is a mechanical way of accumulating those minutes without deciding to. Free, no account, no card. Pro is $4.99 a month for scheduling around your day, custom exercises, ambient audio, stats and leaderboards. The timer is here if you want to start this afternoon.