Analysis of mortality experience amongst Canadian nuclear power industry workers following chronic low-dose exposure to ionizing radiation

The causal link between external ionizing radiation dose and increased risk of a number of cancers has been firmly established in the study of survivors of the atomic bombings in Japan (A-bomb study). Linear extrapolations from this and other high-dose studies have been used to predict risks associated with low-dose low-dose-rate exposures because many individuals are routinely exposed to such doses by virtue of their occupation or residence near radiation facilities. Direct measurements of risks associated with such exposures have been done primarily amongst nuclear workers.; Our cohort included 45,467 nuclear power industry workers from the Canadian National Dose registry monitored for more than one year for chronic low-dose whole-body ionizing radiation exposures from 1952 until the end of 1994 (average follow-up = 14.7 years, mean cumulative dose = 14.3 milliSievert (mSv)). Our study showed that overall mortality was lower than age-, gender, and calendar year-specific Canadian population mortality rates (SMR = 0.57, 95% CI: 0.54, 0.60), which is consistent with a substantial healthy worker effect .; In the first study we found an increased excess relative rate (ERR) of 2.2 per Sv (90% Cl: 0.02, 5.47) for total solid cancer mortality. We also found significantly increased risks for cancers of rectum and lung. Risk estimates for cancers of esophagus, colon, and buccal cavity were positive but non-significant.; In the second study we estimated a sizeable increase in the risk of leukemia excluding chronic lymphocytic leukemia (ERR = 7.79 per Sv, 90% CI: < -1.27, 55.30). Increased risk was also estimated for non-Hodgkin's lymphoma, but not for multiple myeloma, which has been shown to be associated with low-dose radiation in several occupational studies.; The link between exposure to ionizing radiation and mortality from noncancer causes has not been established conclusively. In our third study we estimated an ERR = -0.004 per Sv (90% CI: -1.10, 1.62) for mortality from noncancer diseases excluding accidents. Mortality from cardiovascular and circulatory diseases, which were previously linked to radiation exposure in some occupational studies, was also not associated with dose.; The probabilities of observing our results if the true estimates of association between radiation dose and mortality were equal to the estimates from the A-bomb study are very high (0.22, 0.75 and 0.88 for solid cancers, lymphatic and haemopoietic cancers and all noncancers respectively). Thus, the present results are compatible with those of the A-bomb study. Despite the low power to detect statistically significant associations, our study has provided some evidence that risks extrapolated from high-dose studies do not significantly underestimate risks incurred at low doses and low dose-rates. The combination of the present results with other studies of nuclear workers (currently under way under the coordination of the International Agency for Research on Cancer) would increase the power of the study and will produce more stable estimates.