CONTRACEPTIVE USE AND RISK OF CANCER IN SHANGHAI

描述：(申请人的描述)可能的致癌作用 各种节育方法一直是公共卫生关注的问题， 在过去的30年里，这是许多流行病学研究的主题。 然而，问题仍然存在，对这些问题的答案将在评估中具有价值 各种特定代理和程序的总体风险和收益。 这是一个不寻常的机会，可以使用 前瞻性研究设计由267,000名女性纺织品队列提供 上海的工人被招募参加乳房的随机试验 一九八九至九一年度的自我检讨。所有女性在入境时都接受了面谈 以获得有关他们避孕做法的信息，并 乳腺癌的危险因素。我们建议将计算机化的文件 在上海的两个癌症登记处，以确定哪些女性 患乳腺癌、子宫内膜癌、卵巢癌、子宫颈癌和肝癌 在头七年的后续行动中(180多万妇女年)。 将使用泊松来估计其中一种或多种癌症的比率 与以下一个或多个曝光有关的回归技术： 口服避孕药、注射避孕药、输卵管结扎术、宫内节育器和 人工流产。这是有史以来最大的女性群体之一。 为流行病学目的而组装，以及它在开发中的位置 国家提供了一个机会来评估这些制剂的致癌风险 和感兴趣的程序在不同于其他程序的环境中 关于癌症发生模式的队列研究和 存在潜在的混杂和影响修正的变量。

DESCRIPTION: (Applicant's Description) The possible carcinogenic effects of a variety of birth control methods have been of public health concern and the subject of numerous epidemiologic studies during the past 3 decades. Yet, questions remain, answers to which would be of value in assessing overall risks and benefits of a variety of specific agents and procedures. An unusual opportunity to address a number of specific issues using a prospective study design is afforded by a cohort of 267,000 female textile workers in Shanghai who were recruited into a randomized trial of breast self examination in 1989-91. All women were interviewed in person on entry into the study to obtain information on their contraceptive practices and risk factor for breast cancer. We propose to match the computerized files of this cohort to two cancer registries in Shanghai to identify women who developed cancers of the breast, endometrium, ovary, cervix uteri, and liver during the first seven years of follow up (over 1.8 million women-years). Rate ratios of one or more of these cancers will be estimated, using poisson regression technique in relation to one or more of the following exposures: oral contraceptives, injectable contraceptives, tubal ligation, IUD and induced abortion. This is one of the largest cohorts of women ever assembled for epidemiologic purposes, and its location in a developing country provides an opportunity to assess carcinogenic risks of the agents and procedures of interest in a setting that differs from that of other cohort studies with respect to patterns of cancer occurrence and the presence of potentially confounding and effect modifying variables.