Last year, reported cases of West Nile virus in the United States hit their highest levels in nearly a decade. It’s a good reminder to keep protecting yourself from getting bitten, but it also begs the question: Is this just a sign of a much bigger threat? The answer is just as wily as the pesky mosquito.

According to recent data published June 28 in CDC’s Morbidity and Mortality Weekly Report, the federal public health agency received reports of 5,780 nationally notifiable arboviral disease cases in 2012. (Arboviral diseases are those transmitted by arthropods, such as ticks and mosquitoes.) That year’s multistate outbreak of West Nile virus accounted for 98 percent of the 5,780 cases and represented the highest reported caseload of West Nile virus since 2003. West Nile reports came in from 48 states, Washington, D.C., and Puerto Rico — the virus has yet to be detected in Hawaii and Alaska. Mosquitoes catch the virus by feeding on infected birds and then pass it on to humans.

Of the 5,674 West Nile cases reported in 2012, 2,873 were neuroinvasive, meaning the virus got into the brain (encephalitis) or spinal cord (meningitis), causing serious health problems. But most people who contract West Nile won’t even experience symptoms. The MMWR report states that “extrapolating from the 2,873 West Nile virus neuroinvasive disease cases reported, an estimated 86,000–200,000 non-neuroinvasive disease cases might have occurred in 2012.” The report’s authors — Nicole Lindsey, Jennifer Lehman, Erin Staples and Marc Fischer — go on to write:

Reported numbers of arboviral disease cases vary from year to year. It is not clear why more (West Nile virus) activity occurred this year than in recent years. The weather, numbers of birds that maintain the virus, numbers of mosquitoes that spread the virus, and human behavior are all factors that can influence when and where outbreaks occur. Because of this complex ecology, it is difficult to predict how many cases of disease might occur in the future and in what areas.

So while West Nile virus was first detected in the United States in 1999 in New York and even though it’s had the attention of scientists ever since, the mosquito-borne disease still remains a bit of a mystery.

“Last year’s case numbers are probably just a fraction of what really happened,” said A. Marm Kilpatrick, assistant professor in the Department of Ecology and Evolutionary Biology at the University of California-Santa Cruz.

With declines in federal funding support for surveillance activities, it’s difficult to get a concrete handle on just how much West Nile virus is spreading, Kilpatrick told me. He said that while about 40,000 cases of West Nile have been reported since it first showed up in the U.S., researchers estimate that about 2.5 million people have actually been infected, with the overwhelming majority experiencing no symptoms. (And people who are infected with the virus, whether they develop symptoms or not, are believed to gain lifelong immunity.) What we do know is that so far West Nile virus has infected more than 300 species of birds, more than 60 species of mosquitoes and about 30 mammals — “it can infect just about anything,” Kilpatrick noted.

Beyond case numbers, trying to predict how the virus and its host will move, what communities are vulnerable, and where virus hotspots might pop up are formidable challenges. Because so many environmental and biological factors affect the virus and its spread, forecasting a particular year’s West Nile threat is more a process of educated guessing.

For example, according to a 2012 Lancet article Kilpatrick co-authored on the dynamics and drivers of vector-borne zoonotic diseases, the appearance of such diseases locally is driven by a combination of environmental and social conditions, ranging from  poverty and conflict to temperature and rainfall to land use and development. And even those influences are still the subject of ongoing research, Kilpatrick told me.

For example, he said, one major hypothesis is that the effects of climate and warming temperatures will enable the virus to replicate more quickly inside its mosquito host, which can lead to bigger epidemics among humans. But the research so far isn’t rigorous enough to make a definitive cause-and-effect connection, Kilpatrick said. There’s also the hypothesis that lots of rainfall will lead to more mosquito breeding, which could then lead to more West Nile virus. But even that isn’t for sure, as too little or too much rain could disrupt breeding patterns — Kilpatrick called it the “Goldilocks theory.”

So what if an area experiences a serious epidemic one year — such as last year in Dallas, where the mayor actually declared a state of emergency — is that a solid predictor? Maybe, maybe not. An epidemic one year could mean the local human and bird population built up an immunity to the virus, thwarting the virus the following year, Kilpatrick said. Or there could be enough young birds born that second year with no immunity that West Nile virus makes a repeat visit.

“There are so many pieces to the story,” Kilpatrick said. “The reason I have a job is that figuring out the dominant factor is not a trivial matter.”

In Louisiana, where residents often joke that the mosquito is the official state bird, 155 cases of neuroinvasive West Nile virus were reported in 2012. Gary Balsamo, state public health veterinarian and assistant state epidemiologist with the Louisiana Office of Public Health, said that referring to neuroinvasive numbers is the most accurate way to gauge the ebb and flow of West Nile virus. He told me that even though medical diagnosis of the virus is improving, the 2012 caseload numbers represent a “real increase, so something was going on.”

“Still, we can’t really pinpoint any one factor responsible for that,” said Balsamo, who’s also president-elect of the National Association of State Public Health Veterinarians. “Over the years, people have tried to come up with methods to model (the spread of West Nile virus), but I’m not sure anyone’s done that great of a job simply because the disease is so multifactorial.”

Full of marshes, wetlands and mosquitoes, Louisiana is no stranger to arboviruses, Balsamo noted, and so surveillance and reporting of neuroinvasive disease cases as well as awareness among physicians to test for West Nile is well established. Still, the “best predictor of whether we’ll have cases is the calendar…regardless of what happened the previous year,” he said. In other words, summer is West Nile season so protect yourself from mosquitoes.

“We’ve told people in the past that we don’t have really good predictive models so it’s best to simply avoid mosquito bites,” Balsamo said. “This disease is affected by climate, by rainfall, by terrain…there are so many factors in play here that I don’t know if we’ll ever get really good at predicting it. So whenever you have something like that the best lesson is to always be vigilant.”

To learn more about West Nile virus and how to protect yourself, visit CDC. For a copy of the recent MMWR report, click here.

Kim Krisberg is a freelance public health writer living in Austin, Texas, and has been writing about public health for more than a decade.