Abstract
The aim of this study is to investigate the secular trends in the mean birth weight of the total births (including the single and multiple births) by prefecture and region, using the vital statistics of the Japanese government.
The tables of frequency distribution of the birth weight of total births by sex and by prefecture are available only from the year of 1960 and annually since 1968, with the 10 classes of 500g intervals: ∼999; 1, 000∼1, 499;……; 4, 500∼4, 999; 5, 000∼.
The mean birth weights and their standard deviations of each available year (1960, 1968∼74) were computed by sex and by prefecture and region (Table 1, 2). The remarkable increase of the mean birth weight is readily noted at the prefectural and regional levels (Fig. 1). This probably is due to the socioeconomic improvement of Japan during these periods. The regression coefficients of the straight lines were computed from the time series of the mean birth weight. This linear regression coefficient corresponds to the annual increase rate of each prefecture and region between 1960 and 1974 (Table 1, 2). The statistically significant negative associations are recognized between the annual increase rates and the mean birth weights of certain years (especially of 1960) (Table 3). This negative association suggests that the annual increase rate is closely related to the levels of the birth weight of each prefecture and region. According to the regression equation in Fig. 2-(1), the mean birth weight in 1960 is 3, 251g in males and 3, 165g in females at the zero point of the annual increase rate at the prefectural level. A discussion was made about the terminal level of the mean birth weight in the future.
Then the coefficients of the rank correlation and the coefficients of the variation of the mean birth weight were computed by sex in each year at the prefectural and regional levels. In spite of the gradual decrease of the variation in time, the areal differences of the mean birth weight were clearly noted (Table 4, 5). At the prefectural level for example, Hokkaido and Aomori are always classified in the higher group, Shizuoka and Shimane always in the lower group (Table 6, 7).
In the discussion we emphasized the necessity and the importance of further investigation about the genetic and environmental (natural, socioeconomic and biological) factors related to the areal differences during these periods in Japan, so as to clarify the public health significance of the birth weight as a community health indicator.
