To further understand the risk of stomach cancer after fractionated high-dose radiotherapy, we pooled individual-level data from three recent stomach cancer case-control studies. These studies were nested in cohorts of five-year survivors of first primary Hodgkin lymphoma (HL), testicular cancer (TC) or cervical cancer (CX) from seven countries. Detailed data were abstracted from patient records and radiation doses were reconstructed to the site of the stomach cancer for cases and to the corresponding sites for matched controls. Among 327 cases and 678 controls, mean doses to the stomach were 15.3 Gy, 24.7 Gy and 1.9 Gy, respectively, for Hodgkin lymphoma, testicular cancer and cervical cancer survivors, with an overall mean dose of 10.3 Gy. Risk increased with increasing radiation dose to the stomach cancer site ( P < 0.001) with no evidence of nonlinearity or of a downturn at the highest doses (≥35 Gy). The pooled excess odds ratio per Gy (EOR/Gy) was 0.091 [95% confidence interval (CI): 0.036–0.20] with estimates of 0.049 (95% CI: 0.007–0.16) for Hodgkin lymphoma, 0.27 (95% CI: 0.054–1.44) for testicular cancer and 0.096 (95% CI: –0.002–0.39) for cervical cancer ( P homogeneity = 0.25). The EOR/Gy increased with time since exposure ( P trend = 0.004), with an EOR/Gy of 0.38 (95% CI: 0.12–1.04) for stomach cancer occurring ≥20 years postirradiation corresponding to odds ratios of 4.8 and 10.5 at radiation doses to the stomach of 10 and 25 Gy, respectively. Of 111 stomach cancers occurring ≥20 years after radiotherapy, 63.8 (57%) could be attributed to radiotherapy. Our findings differ from those based on Japanese atomic-bomb survivors, where the overall EOR/Gy was higher and where there was no evidence of an increase with time since exposure. By pooling data from three studies, we demonstrated a clear increase in stomach cancer risk over a wide range of doses from fractionated radiotherapy with the highest risks occurring many years after exposure. These findings highlight the need to directly evaluate the health effects of high-dose fractionated radiotherapy rather than relying on the data of persons exposed at low and moderate acute doses.

Exposure to radioactive iodine ( 131 I) from atmospheric nuclear tests conducted in Nevada in the 1950s may have increased thyroid cancer risks. To investigate the long-term effects of this exposure, we analyzed data on thyroid cancer incidence (18,545 cases) from eight Surveillance, Epidemiology, and End Results (SEER) tumor registries for the period 1973–2004. Excess relative risks (ERR) per gray (Gy) for exposure received before age 15 were estimated by relating age-, birth year-, sex- and county-specific thyroid cancer rates to estimates of cumulative dose to the thyroid that take age into account. The estimated ERR per Gy for dose received before 1 year of age was 1.8 [95% confidence interval (CI), 0.5–3.2]. There was no evidence that this estimate declined with follow-up time or that risk increased with dose received at ages 1–15. These results confirm earlier findings based on less extensive data for the period 1973–1994. The lack of a dose response for those exposed at ages 1–15 is inconsistent with studies of children exposed to external radiation or 131 I from the Chernobyl accident, and results need to be interpreted in light of limitations and biases inherent in ecological studies, including the error in doses and case ascertainment resulting from migration. Nevertheless, the study adds support for an increased risk of thyroid cancer due to fallout, although the data are inadequate to quantify it.

Schafer, D. W. and Gilbert, E. S. Some Statistical Implications of Dose Uncertainty in Radiation Dose–Response Analyses. Radiat. Res. 166, 303–312 (2006). Statistical dose–response analyses in radiation epidemiology can produce misleading results if they fail to account for radiation dose uncertainties. While dosimetries may differ substantially depending on the ways in which the subjects were exposed, the statistical problems typically involve a predominantly linear dose–response curve, multiple sources of uncertainty, and uncertainty magnitudes that are best characterized as proportional rather than additive. We discuss some basic statistical issues in this setting, including the bias and shape distortion induced by classical and Berkson uncertainties, the effect of uncertain dose-prediction model parameters on estimated dose–response curves, and some notes on statistical methods for dose–response estimation in the presence of radiation dose uncertainties.

Combined analyses of data on 260 life-span beagle dogs that inhaled ${}^{238}{\rm PuO}{}_{2}$ at the Inhalation Toxicology Research Institute (ITRI) and at Pacific Northwest National Laboratory (PNNL) were conducted. The hazard functions (age-specific risks) for incidence of lung, bone and liver tumors were modeled as a function of cumulative radiation dose, and estimates of lifetime risks based on the combined data were developed. For lung tumors, linear-quadratic functions provided an adequate fit to the data from both laboratories, and linear functions provided an adequate fit when analyses were restricted to doses less than 20 Gy. The estimated risk coefficients for these functions were significantly larger when based on ITRI data compared to PNNL data, and dosimetry biases are a possible explanation for this difference. There was also evidence that the bone tumor response functions differed for the two laboratories, although these differences occurred primarily at high doses. These functions were clearly nonlinear (even when restricted to average skeletal doses less than 1 Gy), and evidence of radiation-induced bone tumors was found for doses less than 0.5 Gy in both laboratories. Liver tumor risks were similar for the two laboratories, and linear functions provided an adequate fit to these data. Lifetime risk estimates for lung and bone tumors derived from these data had wide confidence intervals, but were consistent with estimates currently used in radiation protection. The dog-based lifetime liver tumor risk estimate was an order of magnitude larger than that used in radiation protection, but the latter also carries large uncertainties. The application of common statistical methodology to data from two studies has allowed the identification of differences in these studies and has provided a basis for common risk estimates based on both data sets.

Using data on 3117 rats exposed by inhalation to radon, radon progeny and uranium ore dust, the hazard function (or age-specific risk) for lung tumor incidence was modeled as a function of exposure, exposure rate and other factors. The overall estimate of lifetime risk was 237 cases per 10 6 rats per WLM (237 per 10 6 WLM), reasonably comparable to estimates obtained from data for humans. The data below 1000 WLM (20-640 WLM) were consistent with linearity with positive excess risks at all levels; however, evidence of statistically significant excess risk was limited to exposures of 80 WLM or greater. Evidence for an inverse exposure-rate effect was limited primarily to cumulative exposures exceeding 1000 WLM (1280-10,240 WLM) and to comparison of results at 100 and 1000 WL. Even at these levels, the possibility that the effect might be explained by time since last exposure or by heterogeneity across experiments could not be entirely excluded. The inverse exposure-rate effect was strongest for epidermoid and adenosquamous tumors, and the only indication of such an effect at exposures below 1000 WLM was modest evidence (P = 0.024) in analyses limited to these tumors. When all lung tumors, or all malignant lung tumors, were included, there was no evidence of such an effect below 1000 WLM. These data support the viewpoint that the inverse exposure-rate effect is primarily a high-dose phenomenon.

Updated analyses of mortality data on workers at the Hanford Site, Oak Ridge National Laboratory (ORNL), and Rocky Flats Weapons Plant are presented with the objective of providing a direct assessment of health risks resulting from protracted low-dose exposure to ionizing radiation. For leukemia, the combined excess relative risk estimate was negative (-1.0 per Sv), and confidence limits excluded risks that were more than slightly larger than those forming the basis of ICRP recommendations. For all cancer except leukemia, the excess relative risk estimate was 0.0 per Sv, but confidence limits indicated consistency with estimates several times those forming the basis of ICRP recommendations. Of 24 cancer types tested, 12 showed positive correlations with radiation dose and 12 showed negative correlations, as would be expected by chance fluctuation. Cancer of the esophagus, cancer of the larynx, and Hodgkin's disease showed statistically significant correlations with radiation dose (P < 0.05), but these correlations were interpreted as likely to have resulted from bias or chance fluctuation. Evidence of an increase in the excess relative risk with increasing age at risk was found for all cancer in both Hanford and ORNL, and both populations showed significant correlations of all cancer with radiation dose among those 75 years and older. Although this age effect may have resulted from bias in the data, its presence suggests that risk estimates based on nuclear worker data be interpreted cautiously.

An important objective of studies of workers exposed occupationally to chronic low doses of ionizing radiation is to provide a direct assessment of health risks resulting from this exposure. This objective is most effectively accomplished by conducting combined analyses that allow evaluation of the totality of evidence from all study populations. In this paper, combined analyses of mortality in workers at the Hanford Site, Oak Ridge National Laboratory, and Rocky Flats Nuclear Weapons Plant are presented. These combined analyses provide no evidence of a correlation between radiation exposure and mortality from all cancer or from leukemia. Of 11 other specific types of cancer analyzed, multiple myeloma was the only cancer found to exhibit a statistically significant correlation with radiation exposure. Estimates of the excess risk of all cancer and of leukemia, based on the combined data, were negative. Upper confidence limits based on the combined data were lower than for any single population, and were similar to estimates obtained from recent analyses of A-bomb survivor data. These results strengthen support for the conclusion that estimates obtained through extrapolation from high-dose data do not seriously underestimate risks of low-dose exposure, but leave open the possibility that extrapolation may overestimate risks.

The dose-response curves for acute radiation symptoms reported by atomic bomb survivors are compared by dose estimation method (the method used to calculate the transmission factor), shielding category, and city. Circular symmetry is also investigated. It is found that response rates for acute symptoms differ considerably by dose estimation method and shielding category even after controlling for both γ and neutron exposure as well as for city, sex, and age at the time of the bomb. One explanation of these results is that the doses of survivors in Japanese type houses estimated by the nine parameter method are subject to less random measurement error, while doses of those survivors who were in the open and shielded by terrain, who were totally shielded by concrete buildings, and who were in factories are subject to especially large random errors. The degree to which systematic bias contributes to these differences could not be determined. These results have important implications for comparisons between cities since Nagasaki includes a far greater proportion of survivors in shielding categories showing weak dose-response relationships than does Hiroshima. The hypothesis that doses might be higher in the westerly direction in Hiroshima is not supported by acute effects analyses, but excess acute effects are found in the north of Hiroshima.

The effects of random dose measurement errors on analyses of atomic bomb survivor data are described and quantified for several analytical procedures. It is found that the ways in which measurement error is most likely to mislead are through downward bias in the estimated regression coefficients and through distortion of the shape of the dose-response curve. The magnitude of the bias and the power for testing the hypothesis of no effect are evaluated for several dose treatments including the use of grouped and ungrouped data, analyses with and without substituting 600 rad for estimated doses exceeding this value, and analyses which exclude doses exceeding 200 rad. The calculations are based on a model in which the error distributions are assumed to be log normal with standard deviations that are 0, 30, and 50%, respectively, of the true dose values. Results are limited to a dose-response function which is linear on total dose. It is found that the commonly applied practice of substituting 600 rad for doses exceeding this value definitely reduces bias in the presence of error. Restricting analyses to doses less than 200 rad reduces bias even more but at the price of considerable loss of power. Both the bias and the power for analyses based on grouped data are very close to the respective bias and power with ungrouped data.

An analysis of an updated file of deaths among Hanford workers is presented. No important changes in the results of the study are noted.

Data from the Hanford plant, where many workers have been employed in jobs involving some exposure to radiation, are analyzed. Mortality from all causes, all cancers, and specific cancer types is related to personnel and exposure data for the population at risk. Results are compared with those of other investigators who have analyzed these data. The mortality of Hanford workers is first compared with that of the United States population and then related to radiation exposure without reference to an outside population. The first analysis shows a substantial "healthy worker effect" and no significantly high standardized mortality ratios for specific disease categories. A test for association of mortality with levels of radiation exposure reveals no correlation for all causes and all cancer. A statistically significant test for trend is obtained for multiple myeloma and cancer of the pancreas but no evidence of a positive correlation is found for 13 other cancer sites including those more typically associated with radiation exposure such as myeloid leukemia and lung cancer. The possibility of other occupational exposures and the lack of reliability with respect to diagnosis of cancer of the pancreas must be considered in interpreting these results. The identified correlations result from a small number of deaths with exposures greater than 15 rem. The lack of correlation for all cancers and for leukemia is by no means inconsistent with current estimates of such effects given the amount of radiation exposure that has been received.