DAY-TO-DAY AND SEASONAL FLUCTUATIONS OF URBAN MORTALITY IN KYOTO, JAPAN

The study and interpretation of temporal variability in mortality requires the consideration of both exogenous and endogenous influences as underlying factors. In the present paper the relative contribution of fluctuations in daily weather was investigated using the unbiased techniques of lagged cross-correlation and spectral analyses. The study focused on patterns of daily mortality in Kyoto, Japan. Studied herein were total mortality of all ages less accidental, ischemic heart (IHD), cerebrovascular (CVD), cardiovascular (IHD + CVD), cancer and among elderly (over 70 years of age) deaths. The meteorological factors were mean, maximum and minimum daily temperature, mean barometric pressure, mean relative humidity, and mean and maximum wind speed. It was found that after extreme weather conditions, such as heat waves (with mean air temperature in excess of 30°C) or the intrusion of cold waves (with mean air temperature below 0°C), mortality increased to about three times the daily average with a lag effect of usually one-three days and up to one week. Over the year, however, weather fluctuations were found to account statistically for no more than 10% of the overall annual variability in mortality. Importantly, the short-term upswings in mortality were usually accompanied by noticeable drops in the number of deaths on the subsequent days suggesting a triggering effect of external factors. The most weather-sensitive mortality group was people over 70 years of age. © 1979 Swets & Zeitlinger B.V.