One of the more disturbing things about parasites is an ability to manipulate their hosts, sometimes to suicidal extremes. The classic example is the liver fluke. It infects an ant as an intermediate carrier, then drives the insect to climb a blade of grass where it is likelier to get eaten by the parasite’s adult-stage host: a cow or other grazing ruminant.

Over the past few years scientists have come to recognize that something along similar lines happens to humans under the influence of one of the deadliest pathogens in our species’s history: The Plasmodium parasite not only causes malaria but also makes victims more attractive to mosquitoes—which then transmit the parasite to still more victims. New research suggests, however, that this situation could actually make it easier to identify and treat carriers who have long eluded medical detection. That could spell the eventual demise of the human Plasmodium parasite, which has killed hundreds of millions of people over the millennia, including 445,000 in 2016.

The new research, published Monday in Proceedings of the National Academy of Sciences, characterizes the distinctive profile of volatile organic compounds (VOCs)—chemicals often perceived as smells—produced by people infected with Plasmodium. Essentially, the researchers identify the odor of malaria as mosquitoes smell it. The scientists propose developing odor-based technology to detect malaria with far greater accuracy than any method currently available, even in patients who show no symptoms.

A team led by Mark Mescher, a behavioral ecologist at the Swiss Federal Institute of Technology Zurich, tested this approach on 400 children at 41 schools in malarial areas near Lake Victoria in western Kenya. The scientists used a portable, briefcase-size device that pulls air from the surface of the skin and collects VOCs in a filter for later laboratory analysis, using gas chromatography–mass spectrometry. The team also took blood samples for testing by two conventional and quicker methods: examination under a microscope or screening for antigens with a rapid diagnostic test. The researchers provided treatment for children who tested positive.

The problem with both of those conventional methods—and even DNA analysis—is they fail to diagnose many cases in which people infected with Plasmodium are partially immune or otherwise show no symptoms. These undetected carriers still have the odor profile of malaria, meaning they are more likely to attract mosquitoes and pass the disease along. And this “hidden reservoir” of infection may account for “up to 90 percent of onward transmission” by mosquitoes, according to the new study. That is a major reason malaria continues to kill so many people, especially children in sub-Saharan Africa. Using the odor profile, on the other hand, “identified asymptomatic infections with 100 percent sensitivity,” the study says, suggesting “significant potential for the development of a robust noninvasive method for detecting malaria infections under field conditions.”

If other studies replicate these findings, malaria would join a growing list of diseases with known odor profiles. These include asthma, tuberculosis, diabetes and numerous cancers as well as diseases of the teeth, gut, heart, liver and kidneys. The new research also comes at a time of increasing commercial and medical interest in developing practical tools for detecting and diagnosing these odor profiles in a timely way. “I’m not sure it’s a huge leap to make this practical in the field,” says study co-author Andrew Read, an evolutionary biologist at The Pennsylvania State University, adding that a breathalyzer-style device might speed detection of hidden carriers and thus hasten the elimination of malaria in areas where it is now endemic. The Bill & Melinda Gates Foundation, which funded the new research, has made not just local elimination but worldwide eradication of malaria one of its stated goals.

Read cautions the new study is based on a single population. “The question is how variable this is around the world,” he says. “Is it the same profile for people with different diets, different lifestyles? We don’t know if we need to fine-tune the profile for different countries.” But, he adds, mosquitoes everywhere seem to cue into something about malaria—“and that suggests it’s the same signal all around the world.”

“It’s really, really interesting” work, says James A. Covington, who works on odor-based medical detection devices at the University of Warwick in England and was not involved in the new research. But he suggests a larger caveat: “The range of chemicals” identified in the new study “are ones that I have seen before for other diseases.” Thus, he says, they might just be the body’s way of saying, “I am ill, I am under stress,” rather than, “I have malaria.” He feels the study would have been more persuasive if it had compared malaria victims with people suffering from another illness, rather than using healthy children as a control group.

Among other scientists who also did not take part in the study, Sabine Dittrich, a public health microbiologist at the Geneva-based Foundation for Innovative New Diagnostics, calls the work an impressive start. “If they have data from 400 children in Kenya, that’s obviously very exciting; 400 is a significant number,” she says. And the noninvasive nature of the test “would obviously make it easier to reach more people,” she notes, adding that this factor is “extremely important” for finding and treating people who are acting as reservoirs of the parasite.

Eric Halsey, a tropical disease physician with the Centers for Disease Control and Prevention’s Malaria Branch, adds one more caveat: Current malaria programs often fail to reach even the obvious symptomatic cases. Having just returned last week from southern Africa, he says the existing diagnosis and treatment technologies are perennially in short supply there, and training programs for health care centers and for community volunteers are inadequate. The World Health Organization reports the $2.7-billion 2016 global investment in the fight against malaria was less than half the $6.5 billion needed to reach targets for malaria control. “Promising research and new tools are always welcome in the malaria field,” Halsey says. But the important thing is to put the tools that are already at hand to work.