Radiation Dose and Solid Cancer Mortality Risk in the Techa River and East Urals Radioactive Trace Cohorts in 1950–2016

The objective of the work was to estimate the dose dependence of mortality risk from solid cancers in a cohort that includes members of two cohorts of residents of the Southern Urals who received chronic environmental low-dose, low-dose-rate radiation exposure from releases of the Mayak Plutonium Production Association. These analyses use dose and dose uncertainty estimates from a recently developed Monte-Carlo dosimetry system. The 47,950 members of the cohort include the Techa River Cohort of people who lived in the villages on the Techa River between 1950 and the end of 1960 and the East Urals Radioactive Trace Cohort of people who lived in territories of Chelyabinsk Oblast contaminated by the explosion of a radioactive waste depository on September 29, 1957, between the date of the accident and the end of 1959. As of the end of 2016, there were 25,723 deaths, including 3,783 solid cancer deaths, with 1,392,394 person years among non-migrant cohort members. The solid cancer mortality rate dose response adjusted for the effect of smoking was estimated using an excess relative risk model. Parameter estimates and confidence intervals were computed using maximum likelihood methods. The corrected information matrix method was used to determine risk estimate confidence intervals (CI) adjusted for dose uncertainty using information on the statistical uncertainty of the parameter estimates and individual dose uncertainty information provided by the dosimetry system. The smoking-adjusted linear excess relative risk (ERR) per 100 mGy for solid cancer mortality was 0.060 (95% CI 0.018 to 0.108) at age 70. The ERR increased significantly in proportion to age to the power 3.1 (95% CI 0.44 to 6.4). The joint effect of radiation and smoking on solid cancer rates appeared to be multiplicative. Adjustment for smoking had little impact on the estimated ERR. Adjusting the ERR confidence interval for dose uncertainty slightly increased the upper confidence bound (adjusted 95% CI 0.018 to 0.120). There was no evidence of nonlinearity in the solid cancer dose response. Except for liver cancer, ERR estimates for various specific types of cancer were positive. However, they were statistically significant only for stomach and female breast cancers. Statistically significant smoking effects were seen for cancers of the lung, stomach, and esophagus. Risk estimates for the two groups in the cohort did not differ significantly. The risk estimates in this cohort were consistent with data in two major occupational cohorts, they were higher than those seen in the Mayak Worker Cohort. While the ERR estimates at age 70 are like those seen in the atomic bomb survivor life span study, the ERR age dependencies were strikingly different. These findings strengthen the evidence for low-dose, low-dose-rate radiation effects on solid cancer mortality rates.