The effect was strongest for close friends but also occurred if friends of friends—or even their friends—gained weight, suggesting that obesity spreads as a kind of social contagion, the same phenomenon popularized in the 2000 book The Tipping Point as an explanation for fads from trucker hats to management philosophies.
Researchers say the contagion effect may occur because we mimic or unconsciously adopt the behavior of people we hold in regard and that it should be taken into account when trying to counteract obesity.
Six in 10 adult Americans are overweight and 31 percent are obese (up sharply from 15 percent 30 years ago), according to a 2004 tally by the Centers for Disease Control and Prevention.
Although another recent study found that a single gene mutation heightened the risk of obesity by up to 67 percent, genetics cannot explain the skyrocketing obesity trend. Lifestyle choices or other features of modern life must be contributing as well. And some researchers have even proposed that infections of gut bacteria account for some cases of obesity.
"It has been fashionable to speak of the obesity epidemic, but we began to wonder if there was really an epidemic: Was there really a person-to-person spread of obesity?" says Nicholas Christakis, professor of sociology at Harvard University and of medical sociology at Harvard Medical School.
Christakis and political scientist James Fowler of the University of California, San Diego, analyzed 32 years' worth of recent records from 12,067 participants in the Framingham Heart Study, which has followed the health of residents of a small Massachusetts town and their offspring every four years since 1948.
Aside from listing their spouses and family members at each follow-up, participants gave names of close friends who would likely know their future whereabouts; more than 70 percent of these were also included in the study, creating a dense social network suitable for identifying epidemiclike effects.
The researchers found clusters of obese friends in the network, but the links did not seem to result from people of similar weight flocking to each other or from undetected causes, they report in the New England Journal of Medicine online.
When two people each listed the other as a friend and one of them packed on the pounds, the second person was 171 percent more likely to become obese. However, if only one member of the pair considered the other a friend, obesity was more likely to spread only to the person holding that view.
A person whose friends had obese friends carried an added 20 percent risk of obesity, which fell to 10 percent for friends of the third degree. In comparison, a chunky sibling increased the risk by 40 percent and a spouse by 37 percent.
Geographic location had no bearing on the results: A portly neighbor had no effect, but a friend who gained weight and lived far away still appeared to raise the risk of obesity. People of the same sex also had a stronger impact on one another.
Christakis says the finding fits with other studies that have shown the benefits of group-based weight-loss programs such as running clubs and Weight Watchers. He adds that the effects of social networks should factor into decisions about antiobesity campaigns costs and potentials for success.
"This is a highly significant study. To my knowledge, there is no other like it," says Barbara Entwisle, professor of sociology at the University of North Carolina and director of the Carolina Population Center.
Health research typically focuses on individual behavior, leaving the role of kin and friends more of a mystery, she says. "I see this as a very significant step in a huge program of research."