Most Americans lose little sleep over dengue fever. The mosquito-borne infection is a leading killer in the tropics and subtropics, but it’s been a long-held belief that ubiquitous air-conditioning, few open windows and limited time outdoors protects us from dengue. And in fact, for the past century most U.S. cases (except those near the Texas–Mexico border) were isolated to immigrants or travelers. In recent years, however, locally acquired cases of the disease have started to appear in pockets of the U.S. Now, researchers fear dengue could be gaining a significant foothold here.

One geographic mystery in particular has forced some epidemiological detective work. Despite that fact that large populations of dengue-carrying mosquitoes are found in certain parts of the U.S., outbreaks have yet to be detected in some of those locations—and scientists are questioning these patterns. Answers could help avert future outbreaks. More than 2.5 billion people—almost 40 percent of the world's population—are now at risk from dengue, and the World Health Organization currently estimates there may be 50 million to 100 million dengue infections worldwide every year.

Tucson, Ariz., and Key West, Fla., each have had sustained populations of Aedes aegypti, a dengue-carrying mosquito, for some 20 years. And yet, whereas the virus rippled through southern Florida, documented cases in Tucson are still unknown. The interest is not just academic—answers could help avert future bouts of the virus. Right now, with no vaccine against the virus, the best defense available is reducing mosquito habitat in areas where dengue fever is common. New clues about why some communities appear dengue-hardened could yield better protective measures.

One research group is now scouring data for new insights. The team conducted door-to-door surveys in the summer of 2010 in Tucson and Key West, hunting for clues about what went wrong in the latter, which in 2009 to 2010 experienced the first dengue outbreak in the continental U.S. since the end of World War II (outside of the Texas–Mexico border). The answer: a big shrug. Both communities had similar figures for screens on windows and central air-conditioning usage, so it was unlikely different numbers of mosquitoes were entering their homes. In both communities similar numbers of people spent at least an hour outdoors most days and reported similar habits in applying mosquito repellant. The researchers talked to 400 households in Key West and 372 in Tucson and found that economic and behavioral factors were similar. So what gives?

A leading hypothesis, says study author Mary Hayden, a behavioral scientist at the National Center for Atmospheric Research in Boulder, Colo., is that in Tucson the climate may be too hot or dry for long-term mosquito survival. Although mosquitoes live to adulthood in both communities, those in Tucson might be dying before the virus can incubate in their bodies. Hayden’s research team plans to test this thinking. “We have just finished our first season of collecting mosquitoes but they are still being processed in the lab,” she says. Whereas public health officials in most cases track the number of reported dengue patients, there is typically no surveillance of mosquitoes carrying the virus. And because it takes four to 10 days for an infected mosquito to be capable of transmitting the virus, Key West might provide more favorable climatic conditions for mosquitoes, she posits.

There may be other clues arising from the differences between the two locales. Dengue fever can be relatively mild (with flulike symptoms, headache, joint pain, fever), so it is possible that people in Tucson may have been less sickened but did not seek care as much as patients in Key West did—or the cases were not verified, Hayden says. Another issue could be health care access—perhaps people in Tucson did not have as many health services available or chose not to take advantage of them.

A similar geographical medical mystery is playing out in Mexico right now. In Nogales, Sonora, on the U.S.–Mexico border, dengue-laden mosquitoes are all-too-common but apparently there’s no dengue present. There, too, the risk is heightened because there are fewer barriers to human contact. For example, people in Sonora are less likely to use air-conditioning than in areas of the U.S., Hayden says.

Overall, dengue is challenging to combat. Bed nets, for example, a key tool when it comes to malaria prevention, are relatively useless against dengue, even though in both cases the vector is mosquitoes. The difference is bite timing. Whereas malarial mosquitoes feed at night (when nets around sleeping space are useful), the dengue-carrying variety typically feed during the two post-dawn hours and the two hours before sunset, Hayden says. As for therapy, it’s basically limited to painkillers. Physicians instruct patients to pop a couple Tylenols (not aspirin because that could exacerbate bleeding, a symptom of dengue).

The problem may soon become more challenging, however. There are four confirmed strains of the virus in the world, but this past year a research group presented findings that suggest the possible presence of a fifth strain in Malaysia. Understanding the absence of dengue in Tucson, especially with so few medical treatment tools currently at hand, would be quite an advance.