病因分值在病例对照研究中的应用──肺癌危险因素的交互效应

对广东省２００例原发性女性肺癌的现患病人作１：１配对病例对照研究，在Ｍａｎｔｅｌ－Ｈａｅｎｓｚｅｌ分层分析和条件ｌｏｇｉｓｔｉｃ回归分析的基础上，探讨女性肺癌危险因素的病因分值以及被动吸烟与其他各种危险因素间的交互效应。多因素条件ｌｏｇｉｓｔｉｃ回归分析结果显示：被动吸烟，厨房通风，喜吃咸食，慢性支气管炎、家庭肿瘤史，肺结核，。服和避孕药是女性肺癌独立的危险因素。

Fraction Analysis of the Involvement of Multiple Risk Factors in the Etiology of Lung Cancer: Risk Factor Interactions in a Case-control Study for Lung Cancer in Females

A 1:1 matched case-control study of etiologic fraction (EF) of female lung cancer was conducted in Guangdong Province. 200 female cases with primary lung cancer were selected and 200female controls matched by age, area of residence and education from five hospitals. Multiple conditional logistic regression analysis turned out that passive smoking from spouse, bad ventilation in kitchen, liking for pickles or salted fish, history of chronic bronchitis, family history of tumour, pulmonary tuberculosis,taking oral contraceptive were the independent risk factors of female lung cancer (OR=2.16-40.55, P< 0.05). EF of the independent risk factors were 0.535, 0.432,0.252,0.124, 0.115, 0.072, 0.069respectively. The interactions between passive smoking from spouse and risk factors which included taking oral contraceptive, family history of tumour, bad ventilation in kitchen, history of pulmonary tuberculosis, liking for pickles or salted fish were analyzed. The result revealed that EF(AxB) (aetiologic fraction attributable to interaction) were 0.848, 0.499,0.479, 0.416,0.346 respectively. The interaction index were 0.906,0.543,0.578, 0.427, 0.440. These findings indicate passive smoking from spouse, particularly those had a history of taking oral contraceptive, increases the risk of female lung cancer.