Scientists who study social networks say they could lead to improved methods for tracking infectious disease.
Social networks may provide the key to early detection of influenza and other contagious outbreaks, according to a new study. A team of researchers who tracked a flu outbreak at Harvard University in late 2009 found that students at the center of social networks contracted the illness sooner than those on the outskirts.
The investigation’s lead authors, James H. Fowler, PhD, and Nicholas A. Christakis, MD, PhD, study the biological implications of social networks and are the authors of Connected: the Surprising Power of Social Networks and How They Shape Our Lives. Together, they have uncovered remarkable patterns by dissecting the intersection of social networks with a variety of behavioral, emotional, and physical health issues—from sleep loss and drug abuse to obesity.
In the most recent study, Drs. Fowler and Christakis randomly contacted 319 undergraduates at Harvard University before the 2009 flu season began. The random group then named a total of 425 friends, who served as the comparison group. Previous research on human social networks has shown the friends of randomly selected individuals to be more "connected," with a higher number of connections and a more central position in the network than those who named them.
As the 2009 pandemic developed, the researchers tracked the health status of both student groups using self-reported data and formal flu diagnoses recorded by clinicians at the school’s health services center. Based on clinical diagnoses, individuals in the friend group contracted the flu approximately 13.9 days earlier than the random group, and 46 days before the epidemic peaked in the population as a whole.
“This sensor method could provide significant additional time to react to epidemics in small or large populations under surveillance,” the authors wrote in a report on the study, which was published September 15, 2010, in the journal PLoS ONE.
A video posted on the journal's blog illustrates the spread of the flu among the individuals studied:
The video is composed of 122 frames, one for each day of the study. Red and yellow are used to indicate infected individuals and friends of infected individuals, respectively. Larger nodes represent individuals with more infected friends. “When you watch the video, you get a sense of flu ‘blooming' among people nearer the center of the network earlier in the course of the epidemic,” the authors explained.
The method used in the experiment could be scaled to predict outbreaks in the larger population. Doing so would provide warnings far in advance of current techniques used by the Centers for Disease Control and Prevention, which usually lag 1 to 2 weeks behind the progression of an epidemic, according to the report.
For other articles in this issue, see: