Healthy worker and healthy survivor effects in relation to the cancer risks of radiation workers.
Healthy worker and healthy survivor effects in relation to the cancer risks of radiation workers.
复制标题
健康工人和健康幸存者的影响与放射工作人员的癌症风险有关。
DOI：
10.1002/ajim.4700170201
复制
发表时间：
1990
影响因子：
3.5
通讯作者：
A. Stewart
中科院分区：
文献类型：
--
作者：
A. Stewart
辐射安全法规中纳入的癌症风险系数基于高剂量效应的线性外推[ICRP，19771]。因此，他们假设，即使高于骨髓损伤的阈值剂量，除了癌症之外，也不存在辐射的后期效应。根据辐射效应研究基金会 (RERF) 的说法，这一假设与原子弹幸存者的死亡经历一致 [Beebe et al., 19771]，而且很明显，RERF 基于幸存者的风险估计比 Mancuso、Stewart 和 Kneale (MSK) 基于工人的估计更受到国际放射防护委员会 (ICRP) 和其他标准制定委员会的重视。这些作者（MSK）风险估计的来源是汉福德的辐射工作人员的一项研究，该研究首先比较死于癌症和其他原因的男性的职业暴露，但最终通过生命表中的回归模型方法将所有受徽章监控的工作人员纳入相对风险分析[Mancuso et al., 1977; Kneale 等人，19811。忽略这项工作的一个借口是，对基本相同数据的其他分析与汉福德队列中几乎没有（如果有的话）额外的癌症死亡是一致的[Darby 和 Reissland，198 I； Gilbert 等人，19891。进一步的原因是，这个群体以及代表核工业其他部门的群体的标准死亡率 (SMR) 远低于标准水平[强烈的“健康工人效应”；参见Polednak 和Frome，1981；切科韦等人，1985；史密斯和道格拉斯，1986；贝拉尔等人，1988； Gilbert 等人，19891。但是“官方”拒绝 MSK 估计值的最强有力的原因是它们与一组非常一致的 RERF 估计值有很大不同 [Beebe 等人，1977 年； Preston 等人，19861。RERF 幸存者队列是在广岛和长崎原子弹爆炸 5 年后组建的，在此日期之后观察到的死亡人数与基于国家统计数据（SMR 分析）或相对风险线性模型的预期之间的比较一直给人留下这样的印象：早期事件没有持久影响，除了癌症之外，辐射也没有后期影响。选择只有一个自由度的风险模型并不令人意外，并且显然是以下结论的结果：到 1950 年 10 月，急性重死亡率期间的任何适应性选择都已消失[Beebe et al., 19781]。然而，Stewart 和 Kneale，
The cancer risk coefficients incorporated in radiation safety regulations are based on linear extrapolation of high dose effects [ICRP, 19771. They therefore assume that, even above the threshold dose for marrow damage, there are no late effects of radiation other than cancer. According to the Radiation Effects Research Foundation (RERF), this assumption is consistent with the mortality experiences of A-bomb survivors [Beebe et al., 19771, and it is obvious that the survivor-based risk estimates of RERF are more highly esteemed by the International Commission on Radiological Protection (ICRP) and other standard setting committees than the worker-based estimates of Mancuso, Stewart, and Kneale (MSK). The source of these authors’(MSK) risk estimates is a study of radiation workers at Hanford, which began by comparing the occupational exposures of men who died from cancer and other causes but eventually included all the badge-monitored workers in a relative risk analysis by the method of regression models in life tables [Mancuso et al., 1977; Kneale et al., 19811. One excuse for ignoring this work is that other analyses of essentially the same data are consistent with there being few, if any, extra cancer deaths in the Hanford cohort [Darby and Reissland, 198 I; Gilbert et al., 19891. A further reason is that this cohort, as well as ones representing other branches of the nuclear industry, have Standard Mortality Ratios (SMRs) that are well below par [strong “healthy worker effect”; see Polednak and Frome, 1981; Checkoway et al., 1985; Smith and Douglas, 1986; Beral et al., 1988; Gilbert et al., 19891. But much the strongest reason for the “official” rejections of the MSK estimates is because they are so different from a remarkably consistent set of RERF estimates [Beebe et al., 1977; Preston et al., 19861. The RERF survivor cohort was assembled 5 years after the bombing of Hiroshima and Nagasaki, and comparisons between observed deaths after this date and expectations based either on national statistics (SMR analysis) or on a linear model of relative risk have constantly left an impression of no lasting effects of earlier events and no late effects of radiation apart from cancer. The choice of a risk model that had only one degree of freedom occasioned no surprise and was clearly the result of concluding that any selection for fitness during the period of acute heavy mortality had worn off by October 1950 [Beebe et al., 19781. However, Stewart and Kneale,