Li H Q, Jin S K, Wu F
Shandong Academy of Medical Sciences, Jinan.
Zhonghua Liu Xing Bing Xue Za Zhi. 1995 Aug;16(4):203-6.
The APC (age-period-cohort analysis) model makes the assumption that the mortality mij in a given age-group and year is the (simple) product of three factors: an age-related factor ai, one period-related factor bj and one (birth) cohort-related factor Ck: mij = eai x ebj x eck. By taking the natural logarithm, a linear model results which can be treated with fairly standard statistics techniques: ln (mij) = ai+bj+ck. The mij data from observation data are used to estimate alpha, beta(a), beta(b), beta(c) by linear regression method. The value of EXP (beta) is to estimate the effect of each variable. The APC model is consistent with carcinogenesis modeling of molecular biology on tumour. The results of APC analysis for cervical cancer, male and female stomach cancer show that the risk factors have been changed. In fact, the risk factors of cervical cancer have been decreased since liberation, and the risk factor of stomach was increased in the first period of sixty's. The analytic method used in this paper will benefit the study on epidemiology and etiology of cancer.
APC（年龄-时期-队列分析）模型假定，给定年龄组和年份的死亡率mij是三个因素的（简单）乘积：一个与年龄相关的因素ai、一个与时期相关的因素bj和一个与（出生）队列相关的因素Ck：mij = eai × ebj × eck。通过取自然对数，得到一个线性模型，该模型可以用相当标准的统计技术进行处理：ln(mij) = ai + bj + ck。利用观测数据中的mij数据，通过线性回归方法估计α、β(a)、β(b)、β(c)。EXP(β)的值用于估计每个变量的效应。APC模型与肿瘤分子生物学的致癌模型一致。宫颈癌、男性和女性胃癌的APC分析结果表明，危险因素发生了变化。事实上，自解放以来宫颈癌的危险因素有所下降，而在60年代初期胃癌的危险因素有所增加。本文所采用的分析方法将有助于癌症流行病学和病因学的研究。
