病例对照研究被广泛应用于流行病学等领域, 通过其获得的病例对照数据不但可以用于寻找疾病的 风险因素, 还能够用于次级分析, 即探究与疾病相关的风险因素之间的关系. 文献中已有的方法多集中于研究次级分析 中的均值回归和分位回归, 而众数作为数据中最有可能出现的值, 既是描述数据中心位置的重要参数, 更是 对均值和分位数的重要补充. 因此文章结合病例对照数据的特征, 提出了一种基于估计方程的众数回归方法 用于次级分析, 同时探讨了估计量的渐近性质. 蒙特卡洛数值模拟结果表明文章的估计方法相比于其他方法有更好的有效性和适用性. 最后利用一个乳腺癌数据集说明了文章所提方法的表现性能.

Case-control study is widely used in epidemiology and other fields. The case-control data from it not only can be used to find risk factors of diseases, but also can be used for secondary analysis, namely, to explore the relationships between risk factors of diseases. The existing methods in literatures mainly focus on mean regression and quantile regression in secondary analysis. As the most likely value in data, mode is not only an important parameter to describe the central position of data, but also an important supplement to mean and quantile. Thus, combining the features of case-control data, a mode regression method based on the estimating equation is proposed. The asymptotic properties of the estimator are also discussed. The Monte Carlo numerical simulation results show that the proposed estimation method is more effective and applicable than other methods. Finally, a breast cancer data set is used to illustrate the performance of the proposed method.