Where the air that brought Kawasaki Disease to Tokyo was hanging out up to 10 days before reaching the island (light brown). Dark brown indicates crops. Fig. 1 from Rodo et al. 2014. Click image for source.

In 2012 I wrote a story for Nature about a strange illness called Kawasaki Disease whose cause has eluded scientists for over 50 years. The diseases causes inflammation of the blood vessels in small children that leads to fever, rashes and reddening, and even coronary aneurysms that can cause heart attacks in the young. Whatever the cause, scientists were startled to discover that the agent seemed to be blowing into Japan, Hawaii, and even the western coast of the United States from somewhere in Asia, implying perhaps that a living infectious organism could survive a trans-Pacific crossing.

Now scientists have pinpointed, they believe, the origin of the KD agent in the northeastern part of China based on an analysis of winds around the annual peak of infection. I had a short news story at Nature this past week that describes the new findings in PNAS, including an inferred very short incubation time for KD that almost certainly rules out an infectious cause. Influenza and norvirus, two of the fastest-acting viruses, have around 35 h average incubation times, while KD seems to take less than 24 hours to make itself known. However, their findings do not rule out the role of a living organism.

That paradox is possible because bacteria and fungi often produce toxins that can cause disease symptoms of their own (botulinim toxin being a famous example). The team tested for an association between inorganic chemical pollutants or pollen and could find none, nor was there much industrial activity in the source region. Instead, the scientists believe it's more likely that a toxin made by a plant, plant-decaying fungus or bacterium, or some other biomolecule is triggering a rapid, destructive immune response in genetically susceptible individuals. “We think it's probably triggered by either a toxin or an antigen,” Burns said, “something the immune system can recognize that comes from something that was once alive.”

Indeed, Kawasaki may be a rheumatic fever-like disease in which multiple exposures to a biological agent inadvertantly fools the immune system into attacking the body's own tissues, she said. That is the mechanism that operates in childhood rheumatic fever, in which Streptococcus bacteria trigger immune-based destruction of the heart valve via an accidental cross-reaction of antibodies to the bacteria with proteins in the valves.

Further evidence supporting the feasiblity of a windborne pathogen came in the form of air samples taken by flying into the wind suspected of triggering Kawasaki Disease. To their surprise, the filters contained an unusually high quantity of the yeast Candida, a fungus. It accounted for over 50% of fungal DNA sequences sampled from the filter and was found repeatedly in suspect air. Though other Candida species can infect humans, none has ever been known to cause disease similar to KD, and the major human pathogen C. albicans was not on the filter. On the other hand, no Candida has ever been found in prior airborne sampling over Asia, the Mediterranean, or the Caribbean.

Other fungi are well known for producing mycotoxins like aflatoxin on crops that cause rapid onset of disease symptoms within a short period after inhalation of toxin-laced spores, consistent with the inferred incubation period and exposure route of the KD agent. However, of the known mycotoxins, none produce KD-like symptoms. On the other hand, experimental exposure to sugar polymers called polysaccharides from Candida in the lab has induced symptoms similar to KD in mice, and can be successfully treated by the same immune-suppressing methods that treat KD in humans.

To follow-up on these findings, scientists need to sample the wind aloft during low-KD times to compare it to peak intensity air and confirm there is less Candida during those times, Burns said. Battisti suggested checking the wind arriving from China on the US West Coast for the same spectrum of microorganisms and elevated Candida signal.

It's possible that the interaction of the agent with clouds or the atmosphere is also playing a role. And it's possible, Burns said, that KD cases in other parts of the world like India or the eastern North America share a common cause but a different source that could be investigated by the same methods.

Whatever the cause, it seems to be coming from the intensively cultivated farmlands of northeastern China where corn, rice, and wheat are grown. Kawasaki Disease only emerged post-World War II in Japan, so Jane Burns, a paediatrics researcher at the University of California, San Diego (UCSD), and director of the Kawasaki Research Center at the UCSD Children's Hospital, speculates that something about Chinese agricultural practices that changed after the war may be involved. “There were a lot of changes in China when Mao took over,” she told me. “This area underwent a huge huge shift in types of crop grown and the intensity of agriculture.”

Stay tuned -- there will no doubt be further intriguing developments in this fascinating medical mystery.


Rod X., Curcoll R., Robinson M., Ballester J., Burns J.C., Cayan D.R., Lipkin W.I., Williams B.L., Couto-Rodriguez M. & Nakamura Y. & (2014). Tropospheric winds from northeastern China carry the etiologic agent of Kawasaki disease from its source to Japan., Proceedings of the National Academy of Sciences of the United States of America, PMID: