A Medico Climatorogical Study in Moderation of Mortality Seasonal Variation

A chronological process of mortality seasonal variation by region in U. S. A

Abstract

After their extensive investigations into the seasonal variation of mortality in various countries in the world, the authoresses have found the concentration of deaths in winter in Japan and the West European countries, on the one hand, and, on the other, the steady moderaration of seasonal variation itself in the United States and the North European countries. In this paper, they attempt to study the latter phenomenon in the United States: namely, they study in detail seasonal variation types by climatic region (Renner), by temperature region based upon seasons (Parkins), and by climatic zone in January.
In the 1960's, seasonal variation is conspicuous in the South (Climatic Region No. 1 and No. 2, see Fig. 2 a) but very slight in the North (No. 4 and No. 5 see Fig. 2 a). In the 1930's and 1910's, variation is bigger than in the 1960's, and the summer death peak is very high in some regions.
Classified by temperature region (see Fig. 2 b), seasonal variation is generally small in the 1960's, particularly in the Pacific region (No. 1), but visibly large in the South, especially Temperature Region No. 5. In the 1930's and 1910's, variation is bigger in all the regions, and there is some upcurve in summer in some regions.
A study by climatic zone in January (see Fig. 2 c) shows that seasonal variation is rather big in the South (Region No. 5 and No. 7) but very slight in the North (No. 8 and No. 9) in the 1960's. In the 1930's and 1910's, variation turns out to be bigger than in the 1960's, and a summer upcurve, together with the winter peak, exist in some regions. For all this, signs of seasonal variation moderation are seen in all the zones.
An investigation into the eight geographic regions (see Fig. 5) reveals that the variation curve of total mortality in the 1960's is generally moderate, though there is an upcurve in winter, while on the other hand it is bimodal (one peak in summer and the other in winter) and less moderate generally in the preceding decades. Classified by color, the variation curve is bimodal for the Whites and the Nonwhites in the 1930's and 1910's, but in the 1960's no upcurve is seen in summer for the Whites but the summer upcurve is conspicuous for the Nonwhites in the southern regions.
Infant mortality in the 1960's varies less visibly than total mortaiity in all the geographic regions except the South. Seasonal variation is marked for the Nonwhites in all the regions, especially Region No. 7 (see Fig. 5) where the coefficient of variation is three times larger for the Nonwhites than for the Whites.
Thus, total and infant mortality now vary less tangibly than decades ago. But the moderation of infant mortality variation is more conspicuous and, moreover, differentials by color prove to be much bigger for infant mortality, particularly for the Nonwhites in the South. An regression analysis also accounts for the moderation of seasonal death variation. The coefficient of regression is smaller in the North than in the South, for infant mortality than for total mortality, and also for the Whites than for the Nonwhites. After all, the recent moderation December M. Momiyama (Sakamoto) and K. Katayama 467 of variation is most predominant for White infants.
The summer upcurve gradually disappears decade after decade and the winter peak gets flatter the recent moderation of death seasonal variation, articularly for infants in the southern regions of the United States, bespeaking the influence of artificial climate on human mortality. It is also
worth mentioning that seasonal variation is conspicuous in the South densely populated by the Nonwhites, especially for Nonwhite infants.