Most human cases of West Nile virus have likely resulted from mosquito species that were not thought to play an important role in spreading the disease, according to a new model of disease transmission in the northeastern United States. The findings could help better focus control efforts. "It's going to be useful for distinguishing which mosquitoes are more dangerous," says Andrew Spielman of Harvard's School of Public Health.
West Nile virus spreads to humans when a mosquito bites an infected bird and then bites a person. In the Northeast, where the disease first appeared in the United States in 1999, several species of mosquitoes are involved. Scientists thought that Culex pipiens and C. restuans bit only birds and that other mosquitoes, such as the common Aedes vexans, were responsible for transmitting the virus from birds to people.
The new work rests heavily on research by Laura Kramer of the New York State Department of Health, Albany. While monitoring the disease, Kramer and colleagues have tested countless mosquitoes for West Nile virus. When Marm Kilpatrick of the Consortium for Conservation Medicine, Wildlife Trust, Palisades, New York, teamed up with Kramer's group and others, they noticed that C. pipiens and C. restuans seemed to have more of a taste for mammal blood than previously thought. "The shocker was that these two bird-biting species also bite people," says Kilpatrick. Genetic analyses by other groups back up that conclusion.
Kilpatrick then devised a model to assess the risk of human infection from various mosquito species. C. pipiens turned out to the most likely source of 80% of cases because of its prevalence in the northeast, its ability to transmit the virus, and its habit of biting mammals as well as birds. That could have implications for control efforts, as C. pipiens lives in wet containers, such as old tires and street drains, while Aedes lays its eggs mainly in wetlands and floodwaters.
"It's a useful and original contribution," says disease ecologist Richard Ostfeld of the Institute of Ecosystem Studies in Millbrook, New York. "It emphasizes how important a small number of generalist vectors can be." Kilpatrick is now further testing the model with mosquito and health data from Colorado and Arizona.