论文部分内容阅读
[目的]介绍多因素调整人群归因危险度(PARP)估计方法及其在队列研究中的应用。[方法]利用2002～2006年建立的上海市男性健康队列(SMHS),在拟合多因素Cox风险比例模型基础上,以男性肺癌为例估计其常见危险因素的PARP。文中分别介绍了PAR在未调整和调整混杂因素情况下的估计方法,以及95%可信区间的计算。数据分析通过SAS程序来实现。[结果]上海市男性健康队列共61500名对象,截止2010年12月31日,共发生肺癌507例。调整其他因素对肺癌影响后,吸烟、酗酒、慢性支气管炎史、肺癌家族史对肺癌的多因素调整人群归因危险度及其95%可信区间分别为64.31%(54.95%～72.13%)、8.46%(3.82%～12.97%)、6.43%(1.13%～11.56%)、3.43%(1.15%～5.62%),其他生活方式因素中水果摄入不足、身体消瘦分别可解释18.33%(0.93%～34.60%)和3.16%(0.42%～5.84%)的上海市区男性肺癌的发生。[结论]上海市区男性肺癌的发生很大程度上归因于吸烟、酗酒、身体消瘦、水果摄入不足等不良生活方式因素。多因素调整人群归因危险度是一个重要的流行病学指标,尤其对指导病因复杂的慢性疾病预防和控制的公共卫生实践非常重要,值得推广应用。
[Objective] To introduce a multi-factor adjusted population attribution risk (PARP) estimation method and its application in cohort study. [Methods] Based on the Shanghai M-Health Cohort (SMHS) established in 2002-2006, on the basis of the multivariate Cox proportional hazard model, we used the case of male lung cancer as an example to estimate PARP of common risk factors. The paper introduces the estimation method of PAR without adjustment and adjustment of confounding factors, and the calculation of 95% confidence interval respectively. Data analysis through the SAS program to achieve. [Results] There were 61500 male health cohorts in Shanghai. As of December 31, 2010, there were 507 cases of lung cancer. After adjustment for other factors affecting lung cancer, the attribution risk of smoking and alcohol abuse, the history of chronic bronchitis and the family history of lung cancer in the population were 64.31% (54.95% -72.13%), 8.46%, 3.43%, 1.43%, 1.43%, 1.43%, 5.62%, 8.46% (3.82% ~ 12.97%), 3.37% ~ 34.60%) and 3.16% (0.42% ~ 5.84%) of the incidence of lung cancer in Shanghai urban areas. [Conclusion] The incidence of lung cancer in urban area of Shanghai is largely attributed to the bad lifestyle factors such as smoking, alcohol abuse, body weight loss and insufficient fruit intake. Multi-factor adjusted population attribution risk is an important epidemiological indicator, especially for the public health practice of guiding the prevention and control of chronic diseases with complicated etiology, which is worthy of popularization and application.