The writers' previous repo r t (1968) has revealed that the seasonal variation of mortality can be classified into the summer concentration type for Egypt, the winter concentration type for Japan and England, and the “deseasonality” type for the United States and the Scandinavian countries. Taking for instance the infant mortality in the City of New York, they attempt in this paper to make an intensive study into the last-named phenomenonthe considerable moderation of seasonal variation or deseasonality and its historical process of formation. (Then is presented a series of models for mortalitytemperature correlation decade by decade in Japan.)
Correlation graphs by the decade for infa n t mortality and temperature in New York, in addition to the convariance, analysis, point out the steady decline of the infant death rate, the gradual disapperance of seasonal differences in mortality, and the formation and levelling-off of nearly linear regression for spring and autumnthe quantitative process of slowing-down in the seasonal variation of mortality or deseasonality. Responsible for all this are, among others, the reduction of deaths from infectious diseases in winter and the favorable effect of central heating. The corresponding correlation graphs for Tokyo present a striking contrast to those for the American city.
Taking for instance Japan's total mortality, the authors h a ve prepared models for mortality-temperature correlation by the decade in the past six decades. From the 1900's through the 1930's, there are two upcurves, one in summer and the other in winter, and the gradual diminution of the summer area of distribution make a contrast with the enlargement of the winter one. In the 1950's, the summer upcurve flattens out, whereas the winter peak gets higher and the winter area becomes larger. It is also noteworthy that the point of intersection for spring and autumn regression curves rises in temperature decade after decade, and that the two curves closely approach each other in the 1960's, without, however, crossing each other.
In conclusion, the causes of such chrono logical changes are studied decade by decade. They differ widely from decade to decade, and in each decade there are specific factors for bulging or flattening. Infectious and old-age diseases are responsible for the bulging upcurve in summer in the 1900's; the arrest of the infectious diseases i n the hot season accounts for the termination of the summer peak, an d two smaller bulges appear in the 1920's and 1930's; and the winter high peak in the 1950's and 1960's is ascribed to the concentration of deaths mainly from apoplexy and other old age diseases.
View full abstract