Emory biologists analyzed transportation data and flu cases from across the United States. The graphic of the U.S. interstate commuter network shows the number of people traveling daily between states for work. Credit: Brooke Bozick.
By Carol Clark
To predict how a seasonal influenza epidemic will spread across the United States, one should focus more on the mobility of people than on their geographic proximity, a new study suggests.
PLOS Pathogens published the analysis of transportation data and flu cases conducted by Emory University biologists. Their results mark the first time genetic patterns for the spread of flu have been detected at the scale of the continental United States.
“We found that the spread of a flu epidemic is somewhat predictable by looking at transportation data, especially ground commuter networks and H1N1,” says Brooke Bozick, who led the study as a graduate student in Emory’s Population Biology, Ecology and Evolution program. “Finding these kinds of patterns is the first step in being able to develop targeted surveillance and control strategies.”
The co-author of the study is Leslie Real, Emory professor of biology and Bozick’s PhD adviser.
One of the fundamental ideas in ecology is isolation by distance: The further apart things are geographically, the more distant they tend to be genetically.
This idea applies to disease ecology in the cases of animals that do not travel far from where they are born. Rabies spread by raccoons, for instance, tends to generate a wave-like pattern of transmission across a geographic space.
People, however, are much more mobile, often traveling by rail, road and air.
The human mobility effect of an epidemic stands out starkly on the global scale. For instance, during the 2003 outbreak of severe acute respiratory syndrome (SARS), airline travel clearly connected cases in people from Asia and Canada.
Map shows an example of how commuting communities can differ from state boundaries. Credit: Brooke Bozick.
The researchers wanted to see if they could detect a correlation to mobility and the genetic structure of seasonal flu cases on a national scale for the United States.
The study tapped Genbank, an online, public repository of genetic flu data, to analyze U.S. cases from 2003 to 2013 for two different subtypes of seasonal flu: H3N2 and H1N1. Transportation data for that decade was drawn from the U.S. Census Bureau and the Bureau of Transportation Statistics, to map out networks of air travel and ground commutes, and the number of people moving along them during the flu season.
The researchers compared genetic distance of the flu subtypes with their geographic distance and the measures of distance defined by airline and commuter transportation networks.
They found some correlations in both subtypes for all the distance metrics used. The correlations were seen a greater proportion of the time, however, when looking at commuter movements and the H1N1 subtype.
“H1N1 tends to be a milder subtype of flu that spreads slower, so that may make it easier to pick up the pattern across shorter-distance commutes,” Bozick says. “We think that a similar pattern for H3N2 may exist. The pattern may just be harder to detect, since H3N2 tends to be more virulent and spread faster, from coast-to-coast.”
The study shows that there are underlying spatial patterns in the genetic data, and that they are dependent on how the “distance” between locations is being measured, she adds.
“Humans can move long distances very rapidly so the idea that geographic proximity is key to determining disease spread doesn’t always hold,” Bozick says. “The patterns we found are likely influenced by states with many commuters, and the identification of these states, as well as network pathways that contribute substantially to influenza spread, is an important next step for epidemiological research.”
Dengue mosquitos hitch rides on Amazon river boats
Human mobility data may help curb epidemics