DCEG investigators and other NCI scientists together with investigators from Kazakhstan and Russia are studying thyroid disease prevalence and radiation dose in a cohort of 2,994 subjects exposed at young ages to radioactive fallout from nuclear weapons testing during 1949-1962 at the Semipalatinsk Nuclear Test Site in northeastern Kazakhstan. Radiation exposure to the thyroid gland was from external radiation sources on the ground and other surfaces, and internally from radioactive iodine taken up by the thyroid gland following ingestion of milk from dairy animals grazing on contaminated pasture. Preliminary point estimates of individual external and internal dose averaged 0.04 Gy (range 0-0.65) and 0.31 Gy (0-9.6), respectively. Ultrasound-detected thyroid nodule prevalence, 18% among males and 39% among females, was significantly and independently associated with both external and internal exposure, whereas prevalence of papillary cancer determined by cytopathology (0.3% in males and 1.1% in females) was not significantly associated with radiation dose.
Current efforts are focused on refining individual dose estimates and characterizing their joint uncertainty structure to correct for possible bias in the dose-response estimates. Individual information collected from a basic questionnaire administered to the study population in 1998 and interpreted on the basis of other, more general information on dietary and agricultural practices are being supplemented by new data obtained in 2007 from focus group interviews of persons with direct knowledge of child rearing and animal husbandry in affected villages. This strategy, designed to encourage accurate memory recall, is being used to derive the distributions on individual consumption rates of milk and dairy products; duration of breast feeding; time spent indoors; fraction of population living in wooden and adobe homes and other important parameters of dosimetry models. The resulting data, combined with physical dosimetry data related to fallout deposition, are used to stochastically generate multiple sets of individual dose estimates and associated dose-response analyses for the entire cohort which, when summarized, yield risk estimates that reflect the uncertainty structure of the individual dose estimates.
For more information, contact Cari Kitahara.