[Clustering Sugi-pollen dispersal patterns for the past 26 years]

Pollinosis caused by the pollen of Sugi (Cryptopmeria japonica) trees is the most significant allergic disease occurring in the spring in Japan. For pollinosis patients and medical staff, it is important to know when the pollen dispersion would reach maximum or when the pollen count would decrease as well as knowing what would the total density of pollen grains be. These sorts of information could be useful for the purpose of disease prevention and deciding on the therapeutic regimen. In this study, we presented the sugi-dispersal patterns and cited several examples of the dispersal pattern.

Airborne pollen grains were collected using a Durham sampler. Total annual pollen counts/cm2 were examined. The sugi-dispersal patterns were classified into several groups by cluster analysis using variables of ten days pollen counts distribution from February to April for the past 26 years. (1987-2012).

The annual pollen count revealed an alternate rhythm consisting of an "on" year (high pollen count) and an "off" year (low pollen count). The results of the cluster analysis showed eleven off-years classified as one group (group 1), and fifteen on-years classified into three groups (groups 2A, 2B, and 2C). The dispersal pattern in group 1 was almost symmetrical with the pollen count rapidly decreasing until late-March. On the other hand, the patterns in group 2 were asymmetrical. In group 2A and 2B a high rate of dispersion was indicated after maximum dispersion, whereas in group 2C the high rate of dispersion was indicated before maximum dispersion. In group 2A, a major dispersion of almost 3000 grains was noted in late-March, and immediately proceeded to the cypress (Chamaecyparis) pollen season without any decrease seen in pollen dispersion. The periods of dispersion of over 10 pollen grains/cm2 per day were 38, 47, 47 and 51 days in groups 1, 2A, 2B and 2C, respectively. That in group 2 was significantly longer than that in group 1, but there was no significant differences between groups 2A, 2B and 2C.

In conclusion, in the dispersal pattern whereby a major dispersion was seen in late-March and proceeded to the cypress pollen season such as in group 2A, patients' symptoms might be prolonged or be more serious. This new concept of dispersal pattern could very well be useful for clinical management of pollinosis.