The study showed that obesity spreads within social networks and that the closer the social connection--even if people live in different households many miles apart--the greater the influence on developing obesity. The study, funded by the National Institute on Aging (NIA), a component of the National Institutes of Health (NIH), is the first to provide a detailed picture of the social networks involved in obesity and could prove useful in developing both clinical and public health interventions for obesity.

The analysis was conducted by Nicholas Christakis, M.D., Ph.D., of Harvard Medical School, and James Fowler, Ph.D., of the University of California, San Diego, using data from the Framingham Heart Study. The Framingham Heart Study is supported by the National Heart, Lung and Blood Institute, another NIH component.

"Nearly one in three American adults--66 million men and women--are obese, which puts them at risk for a number of serious health problems, including type 2 diabetes, heart disease and stroke. With the sharply rising rates of obesity in this country, we need to learn as much as we can about contributing factors. This study describes social network influences that might be an important part of that equation," says NIA Director Richard J. Hodes, M.D.

A sedentary lifestyle and increased consumption of high-calorie foods are critical factors in the steep rise in the prevalence of obesity, the researchers note. But they suggest that a hierarchy of influence exists among family and friends on developing obesity, in which the attitudes, behaviors, and acceptance of obesity also might play an important role.

To explore whether obesity spreads from person to person within social networks, the research team gleaned weight, height and other data from the records of 5,124 Framingham Heart Study participants at up to seven time points between 1971 and 2003. In addition, they analyzed similar information from the Framingham records of these key participants parents, spouses, siblings, children and close friends. Together, these individuals formed a large, intertwined social web totaling 12,067 people. The average age of key participants at the inception of the study was 38 years, with a range of 21 to 70 years.

"We were able to reconstruct a large network of individuals who had been repeatedly weighed over time as part of the Framingham Heart Study, and we could see that as one person gained weight, those around him or her gained weight," says Christakis. "We didn't find that people who were overweight simply flocked together. Rather, we found what seemed to be a spread of obesity and that the likelihood of a person becoming obese depended on the nature of the relationship."

"The rising rate of obesity threatens to reverse the decline in disability in the older population, with major implications for the health care system," says Richard Suzman, Ph.D., director of the NIA's Behavioral and Social Research Program. "This seminal study breaks important new ground in showing how social networks may amplify other factors and help account for the dramatic increase in obesity across the population."

Findings include:

A key participant's chances of becoming obese increased by 57 percent if he or she had a close friend who became obese. In same-sex friendships, a close friend becoming obese increased a key participant's chance of becoming obese by 71 percent. However, no such association was found in opposite-sex friendships.

The perception of friendship also was an important factor. When two people identified each other as close friends, the key participant's risk of becoming obese increased by 171 percent if his or her friend became obese. In contrast, a key participant was not likely to become obese if someone claimed a close friendship with him or her but the key participant did not report the friendship.

Among pairs of siblings, one's becoming obese increased the other's chance of becoming obese by 40 percent. This finding was more marked among same-sex siblings than opposite-sex siblings.

In married couples, one spouse's becoming obese increased the likelihood of the other spouse becoming obese by 37 percent. Husbands and wives appeared to affect each other equally.

Obesity spread across social ties, despite geographic distance from one person to another. Further, social distance--the degree of social separation between two people in the network--appeared to make more of a difference than geographic distance in the spread of behaviors and norms associated with obesity. An immediate neighbor's becoming obese did not affect a person's risk of becoming obese.

Smoking behavior was not associated with the spread of obesity from person to person.

"We identified distinct clusters of obese people within social networks, and the clusters spread about three people deep," Christakis says. "People who were only one degree removed from each other socially, such as siblings or close friends, influenced one another twice as much as people who were two degrees removed from each other."

nih/nia