Introverts, take heart: social butterflies might get invited to more parties, but they’ll probably get sicker quicker. And monitoring the status updates of your classmate with 3,000 Facebook friends could potentially alert you to the leading edge of the next campus flu epidemic.
Nicholas Christakis and James Fowler, of Harvard University and UC San Diego, respectively, have collaborated before on research that examines how social networks shape health, economics, and behavior. Their latest study, examining how friend groups characterize the spread of influenza, appears today in PLoS ONE.
They followed 744 Harvard undergraduate students throughout the course of an on-campus flu outbreak in 2009. The students were divided into two groups. The first was a group of randomly chosen individuals. The other was composed of students that were named by members of the first group as their friends &mdash and were therefore more popular. (It’s a quirk of human behavior that when asked to name their friends, people usually name individuals that are more popular than they are. This is known as the friendship paradox, which can be summed up as: “your friends have more friends than you do”.)
The study found that flu spread faster and peaked sooner within the second group than within the random sampling of students, meaning that people more central to the social network contracted the illness sooner.
“People with many friends get the flu [more soon] than others, making them good bellwethers for what will happen to others in the network,” says Fowler.
He says the method could also be used to track the spread of non-pathogenic phenomena, such as fashions, behaviors, and fads. While social networks in which people are physically interacting are more relevant for examining the spread of a flu epidemic, the evolution of certain memes could be evident through online social networks like Facebook.
(It’s still unclear how one’s position in an online social network correlates with a corresponding position in a real-life social network, something Fowler says he intends to study in the near future. He’s planning on working with Facebook on answering this very question.)
The study from Fowler and Christakis is just one example of how epidemiologists are making use of new media. A paper in Nature last year described a way to track the rise of flu epidemics by studying patterns of health-related search queries on Google. Since the frequency of certain search terms was found to correlate with the number of patients visiting their doctors, the researchers said they could accurately estimate the level of influenza activity that week with a lag of only a day.
But the friend factor adds another layer to the predictive advantages of the model. Awareness of how epidemics spread through dense networks could prove useful in planning strategies to counteract the outbreak of infectious diseases in cities.
“The really nice thing about this method is that you don’t have to know the whole network structure &mdash all you need to do is to pick some people at random and have them name their friends. If you monitor these friends, you’ll be able to know in advance what is likely to happen to the whole network. The friends are our crystal ball,” says Fowler.
Image by Spirit-Fire via Flickr Creative Commons