The Perils of Sitting Too Long: What the Research Shows

If you work at a desk, you probably sit for most of your day. The research on what that does to the body is large and fairly consistent. It is also mostly observational, which matters for how much weight to put on any single number. This post goes through the main studies and says what each one can and can't tell you.

Sitting and earlier death

American Cancer Society cohort (2010). Patel and colleagues followed 123,216 US adults who had no history of cancer, heart attack, stroke or serious lung disease. Women who reported sitting more than six hours a day in their leisure time were 37% more likely to die during follow-up than women who sat under three hours. For men the figure was 18%. The link barely changed after adjusting for physical activity, and it was stronger for cardiovascular deaths than for cancer deaths. American Journal of Epidemiology, 172(4).

Meta-analysis of sedentary time (2015). Biswas and colleagues pooled studies of sedentary time in adults. Across studies, more sedentary time was associated with higher all-cause mortality (pooled hazard ratio about 1.24), and with higher cardiovascular disease, cancer mortality and type 2 diabetes risk. The authors noted that the association was strongest in people with low physical activity, and that results varied a lot between studies. Most of the underlying studies relied on self-reported sitting. Annals of Internal Medicine, 162(2).

It is not only how long, but how long without a break

REGARDS cohort (2017). Diaz and colleagues used accelerometers rather than questionnaires, on 7,985 US adults aged 45 and over. Total sedentary time mattered, and so did how it was accumulated. People whose sitting came in long unbroken stretches had higher mortality than people who sat the same total time in shorter bouts, with the shortest-bout pattern (under about 30 minutes) linked to the lowest risk. The highest-risk group sat more than 13 hours a day, often in bouts of an hour or more. Annals of Internal Medicine, 167(7).

Short-term effects you can measure in hours

Long-term cohorts can't prove cause. Controlled experiments can, over shorter periods.

Dunstan et al. (2012). In a randomized crossover trial, overweight and obese adults sat for five hours after a standard test drink. In one condition they sat without interruption. In the others they stood and walked for two minutes every 20 minutes, at a light or moderate pace. Both walking conditions lowered the blood glucose response compared with unbroken sitting, and insulin responses were about 23% lower. Diabetes Care, 35(5).

A workout after work doesn't fully cancel a day in the chair

Ekelund et al. (2016). This harmonised analysis of data from more than a million people found that high levels of physical activity, about 60 to 75 minutes of moderate activity every day, appeared to remove the extra mortality risk of sitting more than eight hours a day. Lower amounts of activity reduced the risk but did not remove it, and the effect was weaker for people who also watched a lot of television. Activity was self-reported. The Lancet, 388(10051).

That is good news if you can manage an hour a day. Most people with a full desk job can't, which is why the next question is what smaller, more frequent movement can do. We cover that in Why a One-Minute Micro Workout Beats Waiting for the Gym.

What to take from this

• The association between long sitting and worse health is consistent across many studies and methods.
• Most of the evidence is observational. People who sit a lot differ from people who don't in ways that are hard to fully adjust for.
• The experimental evidence (such as the Dunstan trial) shows real short-term metabolic effects from interrupting sitting, but over hours, not years.
• Uninterrupted stretches look worse than the same total broken up.

None of this is medical advice. If you have a health condition, check with your doctor before changing how you exercise.