Oh Sit Down! (or Don't)
Sitting all day: what the research actually says (and what I'd actually do)
Short version: sitting a lot is associated with higher risk, but the size of that risk depends enormously on how much real movement you're also getting elsewhere in your day — and that part is very much within your control.
What the actual research says
The clearest piece of evidence here is a 2023 study pooling device-measured data (not self-reported, actual activity trackers) from nearly 12,000 adults over 50, across four cohorts in Norway, Sweden and the US [1]. Sitting more than 12 hours a day was linked to a 38% higher risk of dying over the follow-up period compared with 8 hours a day — but only among people getting less than about 22 minutes a day of moderate-to-vigorous activity (brisk walking pace or harder). Past roughly 22 minutes a day, that extra risk from heavy sitting effectively disappeared. Each additional 10 minutes of that kind of movement was linked to a 15-35% lower risk, with the bigger benefit showing up in the people who sat the most.
A much larger 2026 analysis backs this up at scale [2]: pooling over 135,000 adults from seven cohorts plus the UK Biobank, the researchers modelled what would happen if people added just 5 extra minutes a day of moderate-to-vigorous activity, or cut sedentary time by 30 minutes a day. Their estimate: those small changes, applied across a population, could plausibly prevent somewhere in the range of 6-10% of deaths from more activity, and roughly 3-7% from less sitting. The activity side of the ledger did more heavy lifting than the sitting side.
Where I'd push back a bit
None of this is a randomised trial where some people were told to sit for years and others weren't — it's observational data, so it can't fully rule out that people who move more are also healthier, less stressed, or better off in other ways that independently affect their lifespan. And those "6-10% of deaths prevented" figures are statistical projections from modelling, not something anyone has actually observed happen. I'd treat the direction of the finding as solid and the exact numbers as a reasonable estimate, not gospel.
There's also a smaller study often cited for "get up and walk every 30 minutes," where breaking up sitting with short walks improved blood sugar and blood pressure [3] — genuinely interesting, but it was 11 people in a lab over a handful of days. Promising, not proof.
What I'd actually do about it:
Treat roughly 20-25 minutes a day of real, "slightly out of breath" movement (brisk walk, bike, stairs, a training session) as close to non-negotiable — it's the single change with the most consistent evidence behind it here.
On days you know you'll be glued to a screen for hours, break it up with a 5-minute walk roughly every 30-60 minutes if you can swing it. It's a small study, but it's cheap and the direction fits the bigger picture.
Don't let a standing desk or a fidgety day fool you into thinking you've solved it — light puttering about only showed a benefit in the most sedentary group, and it's no substitute for the 20+ minutes of real effort above.
Judge your day by your movement minutes, not your sitting hours. If you're consistently getting that daily dose of activity, a desk job isn't the health catastrophe the headlines make it out to be.
References
Sagelv EH, Hopstock LA, Morseth B, et al. Device-measured physical activity, sedentary time, and risk of all-cause mortality: an individual participant data analysis of four prospective cohort studies. British Journal of Sports Medicine. 2023;57(22):1457-1465.
Ekelund U, Tarp J, Ding D, et al. Deaths potentially averted by small changes in physical activity and sedentary time: an individual participant data meta-analysis of prospective cohort studies. The Lancet. 2026; published online 13 January 2026.
Duran AT, Friel CP, Serafini MA, Ensari I, Cheung YK, Diaz KM. Breaking up prolonged sitting to improve cardiometabolic risk: dose-response analysis of a randomized crossover trial. Medicine & Science in Sports & Exercise. 2023;55(5):847-855.
Comments