Purpose

Among the causes of death in Japanese, there are many reports that heart disease, cerebrovascular disease, respiratory disease, and unexpected drowning increase in winter. However, there are few studies on the long-term relationship between outside temperature and mortality. In this study, we analyzed the relationship between mortality statistics and the weather data from 1972 to 2015 (44 years) and explained the seasonal variation of death. The purpose of this study is to clarify the seasonal variation using the results and to use it as an index for healthy and safe architecture and city planning.

Method

We merged the mortality statistics from 1972 to 2015 with AMeDAS weather data to add the daily average outside air temperature on the day of death to each death data. After that, we analyzed (1) CSVM, (2) Death Index, and (3) the relationship between a monthly average outside temperature and monthly mortality rate (MOTMMR).

Result

(1) Calculation of the CSVM in Tokyo, Hokkaido, and the whole country showed that there was a peak of CSVM in 1998 nationwide. Next, when the CSVM for each disease was calculated, it was found that this peak was prominent in respiratory diseases. Comparing the CSVM in European countries and Japanese prefectures, the CSVM was low in cold regions and high in warm regions common to Japan and Europe. (2) In Tokyo, the Death Index in the summer season (from July to September) was high until the 1920s and decreased after the 1930s, but in Hokkaido, the Death Index in the summer season was high until 1940 and began to decline after 1950. In recent years, both Tokyo and Hokkaido have tended to rise in winter and decrease in summer. (3) We found the regression line of MOTMMR has three segments. The low-temperature part has a negative correlation, and the high-temperature part has a positive correlation. The slope of the regression line in the low-temperature was steeper in the southern region. Also, In the regression coefficients of MOTMMR from 1972 to 2011, no significant change was observed in Hokkaido, but the tendencies to approach 0 were found in southern prefectures.

Conclusion

(1) CSVM peaked in 1998, and the tendency was remarkable in respiratory diseases. It may be related to influenza vaccine coverage. In addition, common in Japan and Europe, the CSVM was lower in cold regions than in warm regions, and the heating habit reduced the number of deaths in winter in cold regions. (2) About the Death Index, The Death Index in Tokyo and Hokkaido showed a different tendency until 1950 and showed a similar tendency. That implied the development of water supply and sewerage improve summer death. (3) In many areas other than Hokkaido, such as Tokyo, the fluctuations in the monthly mortality rate due to outside temperatures have improved since around 1980.

From the above, low outside air temperature is related to the rise in mortality in winter, which can be prevented by vaccination, improvement of infrastructure, improvement of building performance, and improvement of heating habits.