Epidemiology of twin births from a climatic point of view.

The twinning rate was analysed using figures taken from statistics for Japan. During the years 1955 to 1959 there were 58 570 twin deliveries out of a total of 9 088 233, a frequency of 6-44 per 1000. Using Winberg's differential method for zygosity estimation, the monozygotic twinning rate was 4-04 per 1000 and the dizygotic was 2-40. Twinning rates in 46 prefectures were also estimated and their correlations with meteorological parameters as well as with the mortality from cerebrovascular disease were examined. The twinning rates increased from a low level in the south west to a high level in the north east of Japan. The tendency was more obvious for the dizygotic than the monozygotic twinning rate. A negative correlation was observed between incidence rates of dizygotic twins and mean air-temperatures throughout the year, and a positive one between incidences of dizygotic twins and mortality rates for vascular lesions affecting the central nervous system in 46 prefectures. A seasonal variation of twinning rates was also observed in Niigata City for statistics of births during the periods 1948-55 and 1963-70. This took the form of a bimodal curve, with high conception rates for twin births in intermediate mild seasons, the spring and autumn and, low rates in the more stressful hot and cold seasons. These relationships between climate and twin births within a single race-group are discussed.