Prenatal exposure to atomic radiation and brain damage

The purpose of this study was threefold: to evaluate the risks to the developing human embryonic and fetal brain of exposure to ionizing radiation using the new DS86 doses; to compare the estimate of risk so derived with those based on the earlier T65DR doses; and to present the evidence bearing on a threshold in the low dose region under the two systems of dosimetry, especially for the data on clinically recognized severe mental retardation (SMR) and seizure. Regarding dose-related SMR, IQ scores, school performance and seizures, there was a high temporal correspondence between the T65DR and DS86 dosimetry systems. A linear no-threshold model with both dosimetry systems also revealed that a significant increase in SMR was observed when the subjects were exposed in the uterus during the periods both 8-15 and 16-25 weeks after fertilization. A threshold in the low dose region was not suggested with the T65DR fetal absorbed doses, but suggested with the DS86 uterine absorbed doses. However, the location or even the existence of a threshold during both periods after fertilization was difficult to demonstrate statistically with the DS86 uterine absorbed doses. When two probable nonradiation-related cases of Down's syndrome were excluded, a threshold with a lower bound was suggested to be observed in the 0.10-0.20 Gy region. Both dosimetries indicated a threshold in the dose-response function for mental retardation in the 16-25 week period, probably within the range from 0.23 to 0.70 Gy. The seizure data provided no persuasive evidence of a threshold during the 8-15 week period after fertilization; the 95% lower bound of the estimate of the threshold included zero. Finally, although the mean IQ scores and the mean school performances in the low dose region were similar to the values in the control group, particularly with doses under 0.10 Gy, evidence for a threshold is not compelling. (N.K.)