肺癌非吸烟性危险因素的条件Logistic回归分析

对广州市区1986年全人群原发性肺癌进行了1：1配对病例对照研究。多因素条件Logistic模型拟合结果表明，男女肺癌的充分病因组成不同。除吸烟外，男性肺癌还与肺部疾患史和家族肿瘤史关系密切，而女性则有所不同。结果还表明新鲜蔬菜可降低肺癌危险性，绿茶也有这种倾向。年龄分层的Logistic回归分析的结果提示，尽管以年龄作为配比因素，仍可存在残余的混杂影响。

The Analysis of Nonsmoking Risk Factors and Primary Lung Cancer with the Model of Conditional Logistic Regression

To assess the impotant roles of nonsmokng risk factors on lung cancer,a matched case-control study based on population in 1986 was perfomed. The asults with multiple conditional logistic regresion model indicated that there was different causal attribution on lung cancer between male and female.Besides smoking, respiratory system diseases history and family tumor history were also important risk factors in male, but it was not so in female.Fresh vegetables and green tea might be protective factors against lung cancer, After stratifing analyses with age, the results showed though age was a matching factor ,it still had some confounding action.