How Much Exercise Balances the Health Risks of Sitting All Day?

Back in 2014, I wrote an article describing sitting as “the new smoking.” In my defense, I intended the phrase to be read in air quotes—even at the time, it seemed a bit hyperbolic. In the years since then, the panic over the health dangers of uninterrupted sitting time has receded. The science hasn’t really changed though. There’s still evidence that being stapled to your office chair all day is bad for you, in ways that even the most rigorous exercise routine might not be able to erase.

It’s that last part that remains controversial: the idea that sitting for ten hours is bad even if you, say, ran for two hours that morning. Over the years, studies have seesawed back and forth on the question of whether sufficient exercise is enough to erase the hazards of excessive sedentary time. In an upcoming issue of the journal Medicine & Science in Sports & Exercise, the two sides square off in a formal point–counterpoint–rebuttal–counter-rebuttal debate.

The claim to be debated is “Physical Activity Fully Offsets the Health Risks of Sedentary Behavior.” Here are some of the key arguments.

Why Exercise Trumps Sitting

The case in favor of exercise was authored by Ulf Ekelund and his colleagues at the Norwegian School of Sports Science and the Norwegian Institute of Public Health in Oslo. The crux of their argument is, essentially, “Check the data.”

Large long-term studies have found that the more hours per day you spend sitting, the more likely you are to develop conditions such as heart disease and, eventually, to die prematurely. The key question is whether eight hours of sitting per day is equally bad for someone who does no exercise, someone who does a bit of exercise, and someone who does a lot of exercise.

Early studies suggested that exercise provided partial but not complete protection from the ravages of sitting. But Ekelund and his colleagues aren’t convinced by the data or the statistical methods of these studies. When they aggregated data from 16 studies with more than a million subjects, they reached a different conclusion, as shown in the following graph:

It’s a complicated figure, but it’s worth taking a moment to parse it. Hazard ratio, on the vertical axis, is an estimate of how likely you are to die during the study period. A hazard ratio of 1.1 basically means that you’re 10 percent more likely to die compared to the reference group.

The subjects are split into four groups of four. The group on the far left is the heavy exercisers: they get more than 35.5 MET-hours per week of physical activity, which works out to 60 to 75 minutes of moderate exercise per day. The group on the far right is the non-exercisers, getting less than 2.5 MET-hours per week, or less than five minutes per day.

Within each group, they’re divided based on how much time they spend sitting, from less than four hours per day to more than eight hours. As you’d expect, the hazard ratio within each exercise group generally increases with sitting time. Check out the non-exercisers on the far right: those who sit the least have a hazard ratio of 1.2, while those who sit the most are up near 1.6. That’s a big difference.

But look more closely at the heavy exercisers on the left. Those who sit less than four hours per day are the reference group, with a hazard ratio of 1.0. But those who sit more aren’t noticeably worse off. Even those sitting more than eight hours a day have a modest hazard ratio of around 1.05 with an error bar that passes through the 1.0 line, meaning that there’s no statistically significant increase in risk from sitting for more than eight hours a day.

QED, according to Ekelund.

Why Exercise Doesn’t Trump Sitting

Strangely enough, the crux of the counterargument, as articulated by a group led by Paddy Dempsey of Deakin University in Australia, is also “Check the data.”

Ekelund’s million-subject meta-analysis, for example, relies on self-reported exercise and sitting data, which is notoriously inaccurate. Using objective data collected by wearable accelerometers produces a different outcome in at least some studies, with the harm of excessive sitting not completely eliminated by lots of exercise.

And even if you look back at Ekelund’s data above, you can’t help thinking that the hazard ratios are trending upward with increased sitting. The effect may not be statistically significant at eight hours, but what about ten? Or twelve?

Dempsey and co. have several other lines of argument. One that I don’t find convincing is “dose realism,” which points out that even if it’s theoretically possible to wipe out ills of sitting with exercise, very few people are actually capable of doing and/or willing to do that much exercise. That may be true, but I’m still interested in knowing whether it’s possible—because if it is, I want to be one of the happy few who does it!

A more convincing point, to me, is the question of physiological plausibility. Or to put it another way, is the problem with sitting merely the absence of exercise? Or does prolonged lack of movement mess up your metabolism in ways that are distinct from the good things that exercise does?

There are various lines of evidence suggesting that the absence of muscle contractions for hours at a time causes damage. For example, the frictional drag of blood rushing through your blood vessels when you move triggers the release of nitric oxide, which helps keep the vessel walls supple and responsive. When you stop moving for hours at a time, the absence of this signal leaves your vessel walls stiffer, which over the long term can predispose you to heart disease.

Another sign that sedentary time directly causes damage is the phenomenon of “exercise resistance.” Normally exercise produces a bunch of immediate metabolic benefits: better insulin sensitivity, higher glucose tolerance, lower triglycerides. But if you’re completely sedentary for a few days, then the measurable metabolic benefits of the exact same workout are blunted.

The Verdict

This is one of those debates where I finished reading each side thinking, “Hmm, this sounds pretty convincing, I guess they’re right.” In other words, there simply isn’t enough data at this point to definitively pick a side.

From a theoretical perspective, I find the “physiological plausibility” argument powerful. Sitting isn’t bad merely because you’re not moving around and thus missing out on an opportunity to get fitter; it’s bad because negative things are happening inside your body.

In that sense, saying that exercise can cancel out the negative effects of sitting is like claiming that eating sufficient broccoli will cancel out the negative effects of ice cream. In reality, no matter how much broccoli you eat, you’re still better off eating less ice cream. (Don’t take this example too literally: I’m with Don Kardong, the Olympic marathoner who famously said: “Without ice cream there would be darkness and chaos.”)

In practice, is this something to panic about? Do you need to set an alarm to do jumping jacks every 20 minutes for the rest of your waking life? Ekelund’s data reassures me that, with sufficient exercise, which seems to be the equivalent of about an hour of moderate activity per day, it’s not that urgent. I have a sit-stand desk, and I generally only raise it a couple of times a day for 10 or 15 minutes, when I notice that my legs are getting stiff. At that point, it actually feels better to stand up for a while, so it’s not a burden.

Of course, the reason that my legs are stiff is usually that I’ve done a particularly long or hard run that morning. Ironically, I find it much easier to accidentally stay glued to my office chair for hours on end when I’m not doing much training. The more sedentary you are, the easier it is to stay sedentary. Maybe that’s a rule of thumb to keep in mind: if getting out of your chair feels like a big hassle, then it’s time to stand up.

For more Sweat Science, sign up for the email newsletter and check out my new book The Explorer’s Gene: Why We Seek Big Challenges, New Flavors, and the Blank Spots on the Map.