X-rays and gamma radiation are well-established carcinogens with substantial evidence of carcinogenicity in humans from epidemiological research. In contrast, there is very little epidemiological research on the human health effects of beta radiation. Even less is understood about the effects of chronic exposure to beta-emitters, such as tritium, a low- energy radioisotope of hydrogen that is produced through nuclear fuel production. Radiation protection standards assign a relative biological effectiveness (RBE) of 1 for beta-emitters compared to gamma-emitters. However, a recent scientific review concluded that the RBE of tritium needs reevaluation. The goal of this proposed study is twofold. First, we propose to estimate the RBE of tritium based on data from the existing literature. The ultimate goal is to apply an accurate RBE for tritium to examine the relationship between chronic, low-dose tritium exposure and cancer among a cohort of nuclear production workers at the Savannah River Site (SRS), the nation's only large- scale tritium production facility. Second, we propose to utilize the information regarding the RBE of tritium to reevaluate the cancer mortality associated with other forms of radiation exposure in conjunction with tritium at the SRS facility. Worker information, including annual tritium and gamma radiation dose, is enumerated for 18,883 workers employed from 1951-1999. Using this data, we propose to establish the distribution of the RBE of tritium with gamma as our reference form of radiation. Recognizing that quantification of a dose-response function for the tritium component of SRS workers will be highly imprecise due to the low magnitudes of tritium doses, we will employ Markov Chain Monte Carlo estimation, a technique that will allow us to establish an a priori distribution based on data from laboratory studies and to test how well an epidemiologic dataset - the SRS cohort data -- fits that distribution. Finally, we will use these estimates to provide more accurate evaluation of the relationship between chronic, low-dose radiation exposure and cancer outcomes. Radiation exposure exists in settings outside of the nuclear fuels industry. Other occupational groups, and the general population, also experience chronic, low-dose radiation exposures. Therefore, the results of the proposed research are relevant for a broad range of occupational and environmental exposures.
There is very little epidemiological evidence concerning the effects of protracted exposures to tritium and its relative biological effectiveness compared to gamma- emitters. The USDOE Savannah River Site is unique in that it was involved in tritium production for nuclear weapons. The goal of this project is to evaluate the association between cancer and estimated tritium dose in this cohort. In addition, we will use Markov Chain Monte Carlo methods to combine current information on the RBE for tritium with our epidemiological data in order to estimate a (posterior) distribution of the RBE for tritium.
| Hamra, Ghassan B; Semelka, Richard C; Burke, Lauren M B et al. (2014) Trends in diagnostic CT among fee-for-service enrollees, 2000-2011. J Am Coll Radiol 11:125-30 |
| Hamra, Ghassan Badri; MacLehose, Richard; Richardson, David et al. (2014) Modelling complex mixtures in epidemiologic analysis: additive versus relative measures for differential effectiveness. Occup Environ Med 71:141-6 |
| Hamra, Ghassan; MacLehose, Richard; Richardson, David (2013) Markov chain Monte Carlo: an introduction for epidemiologists. Int J Epidemiol 42:627-34 |
| Hamra, Ghassan; Richardson, David; Maclehose, Richard et al. (2013) Integrating informative priors from experimental research with Bayesian methods: an example from radiation epidemiology. Epidemiology 24:90-5 |
| Hamra, Ghassan B; MacLehose, Richard F; Cole, Stephen R (2013) Sensitivity analyses for sparse-data problems-using weakly informative bayesian priors. Epidemiology 24:233-9 |
