日本衛生学雑誌
Online ISSN : 1882-6482
Print ISSN : 0021-5082
ISSN-L : 0021-5082
わが国における20年間の平均出生時体重の年次推移
単産,全国・県別,性別,1969-88年
土井 徹華表 宏有曽根 智史松田 晋哉
著者情報
ジャーナル フリー

1993 年 47 巻 6 号 p. 1058-1072

詳細
抄録

Using vital statistics (1981-1988) of the Japanese government, the mean birth weight (MBW) classified by sex was calculated from a birth weight distribution of the single births with 500g intervals for all Japan, prefectures and regions. These MBWs were linked with the MBW of 1969-1980 in our previous reports (Ref. 1)2)). The MBWs of single births for twenty years (1969-1988) were investigated and it was pointed out that the MBWs classified by area (all Japan, prefectures and regions) were on the decrease recently and that the areal differences of the MBW were becoming large.
For all Japan a secular change in the MBW was discussed for the term of a standard deviation, a skewness, a kurtosis, and relative frequencies of less than 2500g (weight class 1), those from 2500g to 3999g (weight class 2) and those of 4000g or more (weight calss 3). These investigations revealed that birth weights of Japan have had a tendency to have a small standard deviation, a negative skewness, and a large kurtosis. Weight class 1 decreased until 1976 and increased after this year. Weight calss 2 was on the increase up to 1988, and weight class 3 increased until 1973 and decreased afterward, as if it were accompanied by the MBW.
For the purpose of confirming the contribution of each weight class to the annual change of the MBW, an analytical method for the weight frequency distribution was proposed as follows. Let M, M', pi and pi' represent the MBW of a given year (year A), the MBW of an another year (year B), the relative frequency of the i-th weight class in year A and that in year B, respectively. Suppose that M (i)" represents the MBW when only the i-th weight class has pi as a relative frequency even in year B. Then, as the result of a fairly simple calculation, the difference between the MBW in year A and that in year B, M'-M, is obtained as follows; M'-M=∑(1-pi')(M'-M(i)"). Therefore, the amount (1-pi') (M'-M(i)") is able to be considered as a share of the i-th weight class in relation to the difference of M'-M. Acordingly, the share proportioned to the summation of the absolute amount, |(1-pi')(M'-M(i)")|/∑|(1-pi')(M'-M(i)")|, may represent the degree of the contribution of the i-th weight class to the difference, M'-M.
Dividing the twenty years into an MBW-increasing period (from 1969 to 1976) and an MBW-decreasing period (from 1976 to 1988), the above analysis was performed in these two periods for all Japan by sex. In both periods the change of the MBW was largely due to the change of the class weighing from 3500g to 3999g, and secondarily to the class weighing from 2500g to 2999g, from the point of view of the proportioned share. It is interesting that the third large proportion fell in the class weighing less than 2500g in the MBW-increasing period and that it fell in the class weighing 4000g or more in the MBW-decreasing period. These results were obtained for both males and females. In order to eliminate the bias from the difference of distribution classified by parity, the primiparous birth weight distributions from 1969 to 1988 of males of all Japan were analyzed. In this analysis the same results were obtained. Consequently, it was suggested that the above results were not due to a difference of distribution classified by parity.
Considering that it is important to reduce the number of low-birth-weight infants in maternal and child health practice and to get expectant mothers maintain a proper weight from the viewpoint of maternal and child health guidance, it seems that the change of the birth weight distribution for the last twenty years represents one of the good results of maternal and child health activity.

著者関連情報
© 日本衛生学会
前の記事
feedback
Top