Got an exercise app on your smart phone? It probably lets you share your exercise data with friends and even with strangers around the world.
Now a study finds that sharing this kind of data may make running contagious. People ran more, ran faster and ran more often if their friends using one of these apps did.
And, more interestingly, they were more likely to slack off just a bit if their connections did.
"We found that exercise is socially contagious," Sinan Aral and Christos Nicolaides of the Massachusetts Institute of Technology wrote in their report, published in Nature Communications.
Researchers have found all sorts of interesting things spread within networks of friends, including happiness and obesity.
Related: Your Happiness Could be Contagious
Sometimes it's hard to figure out if people just hang out with other people who are like them or like the same things, or if they're really influencing one another.
Aral and Nicolaides tried to account for this by factoring in weather.
They looked at data from 1.1 million people who used a social media network — they won't say which one — to post and share their running information over five years. They used an app that tracked their running time, speed and so on and posted it automatically at the end of the run.
"The data contain the daily distance, duration and pace of, as well as calories burned during, runs undertaken by these individuals, as recorded by a suite of digital fitness-tracking devices," the researchers wrote.
"When a run was completed, it was immediately digitally shared with a runner's friends. Runners could not choose which runs they shared but rather comprehensively shared all new running information with their friends upon connecting their device to the platform."
Related: Will the Fat Gene Get You?
They found that for every extra kilometer (about half a mile) run by one person, his or her peers ran about half a kilometer more than usual themselves.
And adding 10 minutes to a run led others to add about five minutes.
"Runners are more influenced by peers whose performance is slightly worse, but not far worse, than their own as well as by those who perform slightly better, but not far better, than they do," Aral and Nicolaides added.
"Men strongly influence men, and women moderately influence both men and women. But men do not influence women at all."
Then they factored in weather, using the assumption that if a runner in Chicago on a rainy day behaved more like his or her running buddy in sunny Boston on the same day, it was more likely because of the influence of the friend.
The findings held.
Why do social networks influence people so much?
"When two people have many mutual friends, there are greater opportunities for social sanctions, reputational consequences for misbehavior and social rewards for positive behaviors," the MIT researchers wrote.
"Mutual friends may therefore provide an added incentive to keep up with running buddies because shirking is widely observed in a set of mutually reinforcing relationships."
This might not hold true for everything, they cautioned.
"Our network sample is reasonably representative of the one in five Americans who owns a wearable device and the over 100 million people who use fitness trackers worldwide," they wrote.
"While this is a substantial and relevant group, they may not represent the average person and peer effects may not operate similarly in the absence of devices that socialize health behaviors."
And there are plenty of runners who don't share their data.
The findings do mean, however, that your annoying friend who posts every single workout online may not just be showing off, but subtly trying to get you to join in.