Skip to main content

Think you're popular? Well, name a friend. It turns out that this person is probably more popular than you, a tendency that scientists might be able to use to predict the spread of disease.

But the popular, like canaries in a coal mine, pay a price: They get flu first, on average two weeks sooner than most others, two experts report in the Public Library of Science journal PLoS ONE.

"Being at the center of the network tends to make you happy but it also exposes you to disease," James Fowler of the University of California, San Diego, who worked on the study, said in a telephone interview.

Story continues below advertisement

Fowler and Nicholas Christakis of Harvard University in Massachusetts said the so-called friendship paradox could be important to working out how a flu pandemic or some other nasty virus is likely to behave.

"This would allow an earlier, more vigorous, and more effective response," Christakis said in a statement.

This is how the friendship paradox works. If a person is asked to name a friend, that friend is statistically likely to be more popular than the original individual.

That is because if people are asked to name a friend or two, they are more likely to choose someone who connects them to others, Fowler says. An example is a party, where most guests would name the host as a friend as opposed to the wallflowers at the fringes of the gathering.

Fowler and Christakis are experts on social networks and have used their methodology to show that obesity, smoking and other behaviors are directly related to a person's friends, and the friends of their friends.

They teamed up again to study the 2009 H1N1 swine flu pandemic among 744 Harvard students.

"Our method goes and picks people at random and then we ask them who their friends are and then we study the friends," Fowler said.

Story continues below advertisement

"We studied the H1N1 pandemic last fall in a small group of students. This friend group -- they got the flu about two weeks earlier than the other groups."

Fowler would like to work with other experts on tracking disease, such as Google Flu Trends, a free service of Google.

Google Flu Trends works on the premise that people's Web searching behavior matches circulating disease. When people feel ill, they look up terms such as "fever" and by watching when and where these search terms become more common, health officials should be able to track disease trends.

Could social networks predict the spread of disease?

It does not yet work perfectly but Fowler thinks there may be a way to get the popular "friends" at the hubs of social networks to volunteer to let their Web searches be monitored in such a database.

"This is one way we could really turn friends into our crystal ball," Fowler said.

Story continues below advertisement

This flu season (typically November to April), a new strain of H3N2, called "Perth flu," is predicted to be the flu to fight. In August, the U.S. Centers for Disease Control and Prevention issued the following alert: "Influenza A (H3N2) virus infections have been recently detected in people in a number of states across the U.S., including two small localized outbreaks."

Health authorities are encouraging people to get a flu shot because of the new strain of H3N2 in the vaccine this year.


Report an error
As of December 20, 2017, we have temporarily removed commenting from our articles as we switch to a new provider. We are behind schedule, but we are still working hard to bring you a new commenting system as soon as possible. If you are looking to give feedback on our new site, please send it along to If you want to write a letter to the editor, please forward to