Lund E
Institutt for samfunnsmedisin, Postuttak, Universitetet i Tromsø, Norway.
Int J Epidemiol. 1990 Sep;19(3):527-31. doi: 10.1093/ije/19.3.527.
The relationship between childbearing and mortality from breast cancer has been studied in a cohort of 822,593 currently married Norwegian women with information on parity from the Census in 1970 and follow-up till 1985. All age groups of parous women showed significant trends of decreasing mortality rates with increasing parity. Nulliparous women had the same mortality rates as uniparous women in all age groups. In a stratified analysis, for age group 45-74 years, the relative risk for uniparous women was between 3.8 and 4.5 dependent on age at first birth compared to women with 8-9 children, age at first birth before the age of 25 and last birth after the age of 30 years. In a multivariate analysis age and parity were stronger risk factors than age at first birth, while age at last birth was nonsignificant. Among women 45-74 years the population attributable risk of breast cancer mortality due to childbearing was 72%, using women with 8-9 children as the reference group. The use of different definitions of reference group for parity is discussed.
在一项对822,593名目前已婚的挪威女性的队列研究中,研究了生育与乳腺癌死亡率之间的关系。这些女性有1970年人口普查中的生育信息,并随访至1985年。所有有生育史的女性年龄组均显示出随着生育数增加死亡率显著下降的趋势。在所有年龄组中,未生育女性的死亡率与只生育过一个孩子的女性相同。在分层分析中,对于45 - 74岁年龄组,与生育8 - 9个孩子、首次生育年龄在25岁之前且末次生育年龄在30岁之后的女性相比,只生育过一个孩子的女性的相对风险在3.8至4.5之间,具体取决于首次生育年龄。在多变量分析中,年龄和生育数是比首次生育年龄更强的风险因素,而末次生育年龄则无显著意义。在45 - 74岁的女性中,以生育8 - 9个孩子的女性作为参照组,生育导致的乳腺癌死亡的人群归因风险为72%。文中还讨论了对生育数参照组的不同定义的使用情况。