This article is our first publication on large-scale dose reconstruction done in a cohort of Belarusian persons exposed to radiation in childhood and adolescence following the Chernobyl accident. Overall, individual thyroid doses due to the main pathways of exposure were estimated for 11,732 study subjects. Intake of 131I was the major pathway of thyroid exposure. Among all Belarusian study subjects, the mean contributions to total thyroid dose from sources of exposure other than 131I intake were estimated to be 2% for intake of short-lived radioiodine isotopes and 132Te, 4.5% for external exposure and 1.5% for ingestion of cesium isotopes. This confirms earlier findings on the predominant role of 131I in the radiation exposure of the thyroid as a result of the Chernobyl accident (12, 28).
We found that thyroid doses from 131I intakes for boys were higher than those for girls due to the larger fraction of milk consumers and higher consumption of milk among boys than that among girls. The same gender-specific differences in thyroid doses from 131I intakes were observed previously in a group of 1,615 children from Belarus and Russia exposed as a result of the Chernobyl accident (29).
The thyroid doses from 131I intakes that are calculated in this study range from 0.54 mGy to 33 Gy, i.e., almost five orders of magnitude. The wide variability in dose reflects the variability in 131I deposition across the country, different consumption habits among study subjects, difference in thyroid mass in persons of different ages and other factors. The scaling factor, which is defined as the ratio of the “measured” 131I activity in the thyroid to the “ecological” 131I activity at the time of measurement (Fig. 2), integrates all steps of the thyroid dose estimation: results of direct thyroid measurement, modeling, and personal interview data. The scaling factor is an indicator of the agreement between the dose estimated using the model and the questionnaire data and the dose derived from the direct thyroid measurement. The closer the scaling factor is to one, the closer the ecological dose is to the instrumental dose. To understand fully the dose assessment, we analyzed the outliers with very low and very high scaling factors. In summary (for details, see Supplementary Material Appendix 2:
dx.doi.org/10.1667/RR3153.1.S2) the possible reasons for obtaining widely different values for the instrumental and the ecological thyroid dose appear to be: (a) incorrect answers provided during the interviews, (b) the assignment of the direct thyroid measurement to a wrong subject, and/or (c) clerical errors in the recording of the result of the direct thyroid measurement. It is worth pointing out that the assumptions were made in the determination of the point dose estimates that the recorded results of the direct thyroid measurements were correct and had always been assigned to the right subjects, and that the answers provided during the interviews were also correct.
Uncertainties in Doses
The wide inter-individual variability of the scaling factors obtained in this study shows that there are large uncertainties in the estimated thyroid doses. The analysis of the very low and of the very large scaling factors (for details, Supplementary Material Appendix 2:
dx.doi.org/10.1667/RR3153.1.S2) indicates that clerical errors may have been made, either in the assignment of the direct thyroid measurement or when recording the result of that direct thyroid measurement. In addition to those errors, which cannot be easily quantified, there are many sources of uncertainty that affect the estimation of the thyroid doses of all subjects. The major sources of uncertainty that are currently under investigation include:
Errors in the 131I activities in thyroids that were derived from direct thyroid measurements. These measured errors arose from device error itself, uncertainties in the estimates of the device’s calibration factors, and uncertainties in evaluation of correction factor that takes into account influence of external and internal contamination of human body on measured exposure rate near the thyroid. These sources of unshared errors are important as measured activity defines the individual dose.
The uncertainties attached to the parameters of the ecological model. Although there are variabilities in the 131I deposition in a given location, the same value of deposition needs to be applied for all persons who resided in that location. The majority of the parameters involved in the ecological model are considered to be shared or subject independent.
The uncertainties attached to the biokinetic models. Obviously, there are variabilities in the thyroid mass and metabolic parameters between individuals. These sources of unshared errors are important because the endpoint of the study is the estimation of individual doses.
The uncertainties attached to the information obtained in 2001–2007 during personal interviews regarding relocation history and individual diet. The reliability of this information is not high as it was collected more than 15 years ago after the accident.
Reliability of The Thyroid Dose Estimates
The thyroid dose estimates due to 131I intakes that have been obtained in this study can be compared to those of a similar study that was conducted in an Ukrainian cohort of 13,215 subjects using the same methodology (29). Although the populations were different, the results were very similar in terms of arithmetic mean thyroid dose (0.68 Gy in Ukraine vs. 0.58 Gy in this study), geometric mean thyroid dose (0.23 Gy in both countries), and ranges of thyroid doses (from 0.0006 to 42 Gy in Ukraine vs. 0.0005 to 33 Gy in this study). The average thyroid doses also are in qualitative agreement with the results presented by UNSCEAR (1) for the entire Gomel Oblast of Belarus, where approximately 80% of the study subjects were exposed to 131I fallout: 0.48 Gy for preschool children, 0.25 Gy for school children and 0.15 Gy for adolescents. The UNSCEAR values are somewhat smaller than the average value obtained for the study subjects (0.58 Gy), probably because UNSCEAR refers to the entire population of the Gomel Oblast and not to the population of its most contaminated raions, as in this study.
The two Chernobyl cohort studies in Belarus and Ukraine make use of the best dose estimation methodology that is currently available and are the only ones in which a doserelated measurement (that is, a direct thyroid measurement) was performed on all subjects. In all other studies, the methodology of dose estimation was based on the direct thyroid measurements that were available for only a fraction of the subjects, while the thyroid doses received by the other subjects are derived from relationships with the 137Cs or 131I deposition densities [e.g., (27, 28, 30)]. The strong reliance on the direct thyroid measurements is due to the fact that these measurements can be considered as the most reliable information for dose assessment purposes, despite associated uncertainties arising from errors in the estimation of the 131I thyroidal content and in the evaluation of the 131I intake function (10).
In summary, although a point estimate of dose was provided for each study subject according to the best methodology currently available, there are obvious uncer uncertainties tainties associated with reconstructed doses arising from errors in estimates of 131I activity in thyroid; errors attached to the parameters of the ecological and the biokinetic models; and reliability of the information obtained during interviews regarding personal behavior more than 15 years ago. Calculation for each study subject of a set of 1,000 stochastic thyroid dose estimates which takes into account classification of errors as classical/Berkson and shared/ unshared is underway. For each subject, the resulting database will include, in addition to the 1,000 stochastic thyroid dose estimates, the value of the scaling factor to provide an indication on the quality of the dose. These sets of dose estimates are being used to evaluate radiation risk that takes into account the structure of the errors in the dose estimates.
Radiat Res. 2013 May; 179(5): 597–609.