Revision 1
March 19, 1998
by Radiation, Science, and Health, Inc.,
Edited by J. Muckerheide

1.2.2
Ocupational

"A total of 35,933 white male workers (5,546 deaths) from three United States nuclear weapons plants with lifetime exposures of 2 to 20 cSv had lower total cancer mortality rates than internal controls, p<0.001 (Figure 15) (Gilbert et al., 1989). The continuously decreased rate, shown in the cumulative curve, and the fact that those exposed to 25 cSv had less cancer mortality than those exposed to 13 cSv, p<0.05, strongly suggest that the optimum lifetime exposure for decreased cancer mortality is greater than 25 cSv. Since the follow-up period averaged 19 years, the optimum exposure appeared to be more than 1 cSv per year. The combined workers had a lower cancer mortality rate than that of the United States population; the SMR was 0.79.

Figure 15

"When the leukemia mortality rate of exposed workers from three United States nuclear weapons plants were compared with that of unexposed workers, no statistically significant differences were found (Gilbert et al., 1989). When compared with the respective local populations, the mean SMR for leukemia mortality in all male workers of the three plants was 0.92.

"Cumulative lung cancer mortality of male workers in three United States nuclear weapons plants appeared to decrease as the dose increased (Figure 16) (Gilbert et al, 1989). Only in those with lifetime exposures of greater than 20 cSv was the decrease statistically significant, p<0.001. The SMR for lung cancer mortality in all workers was 0.76."

Figure 16

"During 20 years in a Canadian energy plant, 4,000 nuclear workers with an average exposure of 70 mSv had a lower cancer mortality rate than 21,000 unexposed workers, p<0.001 (Figure 17) (Abbatt et al, 1983). The cancer mortality rate of thermal workers in the plant was comparable with that of the general population of Ontario; the SMR was 0.97. There were no leukemia deaths in exposed workers during this study."

Figure 17

"Deaths in another Canadian energy plant were followed from 1956 to 1985 (Gribbin et al, 1993). A comparison of over 4,000 exposed workers with 4,000 other workers in the same plant showed no significant differences in cancer mortality rates. The authors made an age adjustment without giving enough data to make an age correction. Comparison of all cancer deaths in all workers with the general population gave the following SMRs: all cancer, 0.77; prostate, 1.21; alimentary, 1.02; leukemia, 0.62; and lung, 0.86. None of these differences were statistically significant."

"A study of 95,000 predominantly male workers in several British nuclear weapons plants from 1955 to 1988 involved 6,660 deaths; only 2.7% of the deaths were female (Kendall et al, 1992). The total cancer mortality rate decreased inversely with exposure, p<0.001 (Figure 18). Since workers exposed to a mean of 7 cSv had a lower cancer mortality rate than those exposed to 2.4 cSv and had about the same rate as those who received 25 cSv, the optimum lifetime exposure for the 33 years appears to be at least 20 cSv, about 0.6 cSv per year. When compared with the population of England and Wales, the SMR for all cancer deaths in nuclear weapons plants was 0.86, p<0.001."

Figure 18 .

"Leukemia mortality followed the pattern of total cancer mortality in the British study (Figure 19). Leukemia mortality in exposed workers was less than that of unexposed controls in the same plants, p < 0.001. The optimum appeared to be 10-30 cSv per 33 years. When compared with the general population, the SMR for leukemia mortality in all workers was 0.91, p = NS."

Figure 19

"Conclusions about chronic exposure of humans to low-dose irradiation are based upon almost eight million person-years (Table 2). These studies predominate over those studies which report increased cancer mortality in small pockets of workers, which are included in the larger surveys. Internal comparison with control and exposed workers in the same plant give irrefutable evidence that low-dose irradiation is beneficial. The "healthy worker effect" cannot account for the decreased cancer mortality rates in nuclear workers. The consistently decreased cancer mortality rates of exposed nuclear workers when compared with unexposed workers in the same plants are compelling evidence that the differences observed are not due to a "healthy worker effect." Both groups entered the plants under the same conditions and received comparable medical care. When compared with the general population, the longer average life span of workers should result in a higher cancer mortality rate. It does not. Thus, the "healthy worker effect" helps to validate radiation hormesis in cancer mortality."

TABLE 2. Major Studies of Cancer Mortality in Nuclear Workers