Velema J P, Walker A M
Int J Epidemiol. 1987 Jun;16(2):177-83. doi: 10.1093/ije/16.2.177.
A multiplicative model was fitted to incidence rates of nervous system tumours cross-tabulated for each sex by five-year age groups between ages 35-64 and 51 populations in which at least ten cases had been observed and 60% of cases histologically verified. There was no significant departure from a model in which the age curve was assumed to have the same shape but a different level in each population. On log-log scale this shape was a straight line with identical slope (2.6) for males and females. Population-specific multipliers were highest in Israel and lowest in Asia. Within the US, rates were significantly lower among blacks, Hispanics, Chinese and Japanese compared to whites. A model for both sexes combined showed that sex ratios were rarely below unity but varied significantly between populations. Exposure (or susceptibility) to as yet unidentified aetiological factors appears to be higher in males and whites than in females and non-whites.
采用乘法模型对35至64岁的5年年龄组按性别交叉制表的神经系统肿瘤发病率进行拟合，数据来自至少观察到10例病例且60%的病例经组织学验证的51个人口群体。假设年龄曲线在每个群体中具有相同形状但水平不同的模型与实际情况无显著偏差。在对数-对数尺度上，该形状对于男性和女性而言是一条斜率相同（2.6）的直线。特定人群乘数在以色列最高，在亚洲最低。在美国境内，黑人、西班牙裔、华裔和日裔的发病率与白人相比显著较低。针对男女合并的模型显示，性别比很少低于1，但在不同人群之间差异显著。男性和白人对尚未确定的病因因素的暴露（或易感性）似乎高于女性和非白人。
