The relationship between monthly death rates from pneumonia and bronchitis (P&B), and monthly means of outdoor temperatures, relative humidities, precipitations were observed comparatively with the multiple regression analysis in the regions, Hokkaido, Tokyo, Wakayama, Okinawa prefectures and Kyoto city. Previously reported “Excess Death Index” which is defined as the difference between the death rate calculated with regres-sion curve of annual death rate during 15 years (1977-1991) and the observed actual death rate was used again. The results are as follows: (1) Independent variable in regression is only one, that is month-ly mean temperature, and the death indices (normalized) were found to rise significantly (p<0.05) with descending mean temperature in 5 regions. In addition, in higher temperature, that is in summer, the death indices appeared to rise with rising temperature. (2) Then, from observation of spline curves based upon averages of death indices and monthly mean temperatures for 15 years, higher temper-ature in summer was found to be a risk factor as well as lower temperature in winter. (3) The death index-minimum temperatures in temperature-rising season was obtained: 15.0 for Hokkaido, 22.0 for Tokyo, 23.0 for Kyoto C. 22.5 for Wakayama and 23.5°C for Okinawa. The obvious shift of dependence of death index upon temperature was observed in Hokkaido.
View full abstract