A combination of cross correlation and trend analyses reveals that Kawasaki disease is a pollen-induced delayed-type hyper-sensitivity disease

Int J Environ Res Public Health. 2014 Mar 4;11(3):2628-41. doi: 10.3390/ijerph110302628.


Based on ecological analyses we proposed in 2003 the relation of Kawasaki Disease (KD) onset causing acute febrile systemic vasculitis, and pollen exposure. This study was aimed at investigating the correlation between pollen release and the change in the numbers of KD patients from 1991 to 2002 in Kanagawa, Japan. Short-term changes in the number of KD patients and medium- to long-term trends were analyzed separately. Short-term changes in the number of KD patients showed a significant positive cross correlation (CC) with 9- to 10-month delay following pollen releases, and a smaller but significant CC with 3- to 4-month delay. Further, a temporal relationship revealed by positive CC distribution showed that pollen release preceded KD development, suggesting that pollen release leads to KD development. A trend in patient numbers was fitted by an exponential curve with the time constant of 0.005494. We hypothesized that the trend was caused by the cumulative effects of pollen exposure for elapsed months on patients who may develop KD. By comparing the time constants of fitted exponential curve for each pollen accumulation period with 0.005494, the exposure period was estimated to be 21.4 months, which explains why approximately 50% of patients developed KD within 24 months from birth.

MeSH terms

  • Humans
  • Hypersensitivity, Delayed / epidemiology
  • Hypersensitivity, Delayed / etiology*
  • Japan / epidemiology
  • Mucocutaneous Lymph Node Syndrome / epidemiology
  • Mucocutaneous Lymph Node Syndrome / immunology*
  • Pollen / immunology*
  • Statistics as Topic