HL20: Nicholas Christakis—Connecting Human Networks, Human Health

Most of the network research takes place at The Christakis Lab at Harvard University in Boston and involves elaborate computer modeling.

In one of the lab's best known studies, data from the famous Framingham Heart Study was used to demonstrate how weight gain in one person might ripple though a social network. In looking at the relationships of 12,000 people, Christakis and Fowler found that a person's chances of becoming obese increase by 57% when a friend becomes obese; by 40% if a sibling becomes obese; and by 37% if a spouse becomes obese.

"When people around you gain weight, your attitude about what constitutes an acceptable body size can change, and you might gain weight, too," Christakis explains.

One of his favorite studies involves a new way to predict epidemics. It uses the friendship paradox (in 25 words or less: A given person's friends are probably more popular than that person) to identify who's at the center of a social network and who's at the margins. Think of the people in the center as trendsetters or, as Christakis likes to call them, "early warning systems." Just as this group might be early adopters for fashion, technology, music, etc., they also get sick first and the illness spreads through the entire network.

Christakis says the overarching idea of his social network research is that "people are connected, so their health is connected." Your individual health depends not only on your own choices and behaviors but also on the people who surround you, including people you know and people you don't know.