【摘要】 用Logistic回归进行流行病学流行因素的多变量分析。本文报告了100例乳腺癌病例和200名对照(一半来自医院其他肿瘤病人;另一半来自里委邻居)。用条件Logistic平衡了各种因素(共19个因素)。结果表明:初潮年龄愈早可能发生乳腺癌危险愈大;母乳哺乳可能有保护作用;高脂肪摄入、乳腺疾病和染发都是乳腺癌的危险因素。更多还原

【Abstract】 The Logistic Regression Model (LRM) analysing multiple variables simultaneously is one of the important methods in multivariate analysis of an epidemiological study. One hundred cases of breast cancer patients and two hundred matched controls (50 % comprised of cancer cases from other hospitals, the rest were cases from the same district as the breast cancer cases) were analyzed. The conditional LRM was used and balanced with nineteen indices respectively. Results showed that the earlier the age of menarche the higher the risk might be. Lactation had some effect to protect against breast cancer. High fat diets, breast lesions and usage of hair dye also played risk roles in breast cancer. It is suggested that reasonable diets for woman, surveillance of usage of hair dye, and breast feeding might be beneficial for breast cancer control in Shanghai.更多还原