Though not completely surprising, that finding could be important as parents, after-school teachers and camp counselors try to encourage youngsters to move more and head-off obesity before it starts, researchers said.

The results are also in line with research that's been done in teens and adults, who tend to look like the rest of their friend group in terms of weight and fitness level.

"This is more evidence that peers and social networks do influence health behaviors," said Dr. Pooja Tandon, a childhood obesity researcher from the University of Washington in Seattle who wasn't involved in the new study.

"The next steps will be (understanding) how to harness the power of social networks to promote health behaviors," such as physical activity in kids, she told Reuters Health.

Researchers from Vanderbilt University in Nashville studied 81 racially-diverse public school students, ages five to 12, who went to after-school programs at one of two different sites.

To see how the kids' friendships affected their physical activity -- and vice versa -- pediatrics researcher Sabina Gesell and her colleagues spent time with the students during three week-long periods over the spring of 2010.

During each visit, they asked kids individually who they were friends with in the after-school program. Then, Gesell's team outfitted the youngsters with accelerometers -- small devices that clip on to the belt and measure how active people are at any given time.

Based on accelerometer readings, the students spent an average of 30 percent of their free time at after-school in what the researchers counted as moderate-to-vigorous physical activity, including running around or playing active games.

Over the course of the study, Gesell and her colleagues found, kids didn't make or break friendships based on how active they were compared to other students. For example, those who spent most of after-school running around were equally likely to befriend their active or non-active peers.

Instead, when kids made new friends who were more or less active, they tended to change their own activity level accordingly, the researchers reported Monday in the journal Pediatrics.

"Kids are constantly adjusting their activity levels to match their friends," Gesell told Reuters Health.

That finding starts to hint at possible ways to address evidence that kids are spending more time on the couch and less time running around than in the past -- and becoming overweight and obese as a result.

"If (counselors) can basically use different strategies to encourage activity even in some children, that could have a ripple effect," said Tandon, who also wondered if similar techniques to promote exercise would work during school recess or at home.

Of course, friends in the study didn't always have a positive influence.

"Some kids' activity levels got pulled up by their immediate friends, and others got pulled down," Gesell told Reuters Health.

The question now, she said, is how to "leverage" that finding to encourage less-active kids to get more exercise, and not the other way around.

The research is still a few steps away from leading to changes in how after-school programs are run in the real world, Gesell said. But in the future, counselors could shake up sedentary friend groups and encourage a couple of less-active kids to join those that go straight to the gym or the playground, she added.

"The after-school programs have had this long history of keeping kids safe and keeping them off the street," Gesell said. "Now the thought is, what if we use this ideal arena to improve health?"

SOURCE: bit.ly/jsoh2P Pediatrics, online May 28, 2012.

• A Permanent Economic Emergency