阳江高本底地区恶性肿瘤死亡调查1987～1990年资料的初步分析

Preliminary analysis of data(1987-1990) from investigation of cancer mortality in high background radiation area of Yangjiang, China.

阳江高本底地区恶性肿瘤调查始于１９７２年，目的在于探索该天然放射性高本底辐射连续照射对人群恶性肿瘤死亡率的影响，为小剂量电离辐射致癌危害概率估计和预测提供对人类的直接观察资料。方法本阶段采用定群队列观察恶性肿瘤死亡资料收集分人口学调查和死因确认两步骤。相对危险（ＲＲ）的计算，应用Ｅｐｉｃｕｒｅ（ＨｉｒｏｓｏｆｔＩｎｔｅｒｎａｔｉｏｎａｌＣｏｒｐ１９８８～１９９２）中ＡＭＦＩＴ程序。结果该定群队列高本底地区高、中、低三个剂量组和选自对照地区的对照组共１０６５１７人。１９８７～１９９０年累积观察了４２１６４０人年，期间恶性肿瘤死亡２３１例。各剂量组与对照组比较经性别、年龄组调整的ＲＲ，对全部恶性肿瘤和全部实体癌，除低剂量组外，ＲＲ均＜１。就部位别肿瘤而论，所分析肺癌、肝癌、胃癌、鼻咽癌和白血病中，除鼻咽癌各组和白血病高剂量组外，其他ＲＲ均＜１。但所述各ＲＲ的９０％置信区间较宽，且绝大部分无统计显著性。与以往研究合并资料（１９７９～１９９０）分析，也得到类似的结果。结论综观本阶段研究资料初步分析，可看到高本底地区恶性肿瘤死亡率低于对照的趋势。这重复了以往研究（１９８６年前动态队列研究）的结果。但要得出有统计学意义?

bjective\ The cancer mortality study conducted in the high background radiation area (HBRA) of Yangjiang, China, was started in 1972. The major objective of the study was to estimate cancer risk associated with the low level radiation exposure in HBRA. Since 1991, the cancer mortality data were collected from a fixed cohort instead of a dynamic one as in the previous stage. The purpose of the study in the current stage was to accumulate further person years (Pyr) of observation for improving the statistical power of test in the estimates of cancer risk associated with the natural radiation exposure and to see the reproducibility of the previous results. Methods\ The collection of cancer mortality data consisted of two phases, the survey of demography and the ascertainment of death causes. The estimates of relative risk (RR) were calculated by AMFIT in Epicure (Hirosoft International Corp., 1988—1992). Results\ During the period of 1987—1990 there were 231 cancer deaths among 421640 Pyr at risk in the cohort of 106517 subjects, which were classified into four groups based on the hamlet specific average annual gamma ray absorbed doses: three groups (high, medial and low doses) from HBRA and control group from CA. The RRs adjusted for sex and age for each dose group in HBRA compared with the control group for overall cancers and for all cancers except leukemia were less than 1 except for the low dose group, and there seemed to be a trend of RR decrease with the increase of exposure dose from natural radiation, though there was no statistically significant difference. As for the site specific cancer studied, the cancers of lungs, liver, stomach, nasopharnyx and leukemia, the RRs were also less than 1 except for nasopharnyx cancer and for leukemia in the high dose group. The analysis of combined data (1979—1990) showed similar results. Conclusion\ Although the sample size was not large enough to make definite conclusion statistically, the observation of current follow up period reproduced the previous results obtained from the data up to 1986 that the cancer mortality in HBRA is generally lower than that in CA. Further follow up is necessary.