北京市城区结肠癌发病率分布滞后预测模型研究  

Study on the Distributed Lag Model for Incidence of Colon Cancer of Beijing Urban Area

作　　者：沈默[1] 曲宸绪[2] 游伟程[2] 张联[2] 王涛[1] 

机构地区：[1]北京大学公共卫生学院流行病与卫生统计学系 [2]北京大学临床肿瘤医学院

出　　处：《首都医科大学学报》2008年第4期501-504,共4页Journal of Capital Medical University

摘　　要：目的研究北京市城区红肉购买量对结肠癌发病率的分布滞后效应,构建结肠癌发病率的分布滞后预测模型。方法利用北京城区近24年间结肠癌发病率,以及城区近28年红肉年人均购买量,利用Eviews软件建立结肠癌预测分布滞后模型,分析模型主要指标,检验模型参数。结果人均红肉购买量对结肠癌发病率的影响在5年内不显著(P<0.01),5年后逐渐增大,模型中最大效应出现在14年,模型的R2=0.9851,6年回代误差率均值为2.635%,误差方差为0.498。结论分布滞后模型应用于结肠癌预测时解释性好,预测精度高,红肉消费对人群结肠癌作用5年以后开始增加,最大效应14年时出现。Objective To investigate the distributed-lag effect of red meat consumption on the incidence of colon cancer by establishing distributed lag model in Beijing urban area. Methods Incidence of colon cancer in Beijing urban area during 1982 - 2005 obtained from the Tumor Registration Office and the consumption amount of red meat per year per person in Beijing urban area during 1978 - 2006 were collected from published statistical annual books. The distributed lag model of incidence of colon cancer was established with these data and the incidence rate could be estimated for the future. Results The amount of red meat per year per person is related to the incidence of colon cancer with a lag period of 14 years. Conclusion The distributed lag model with risk factor is an effective model to estimate the tendency of incidence of colon cancer. The consumption amount of red meat affects the incidence of colon cancer in Beijing urban area started from 5 and the maximal effect became obvious in 14 years.

关 键 词：结肠癌 分布滞后模型 分布滞后效应 

分 类 号：R73-31[医药卫生—肿瘤] R181.23[医药卫生—临床医学]