exposure. The number of persons in the Rongelap group is small and many werechildren at the time of exposure, thereby introducing sample size and age factor into the analysis. Finally, variations in the thyroid-absorbed radiation dose were primarily dependent on ageat exposure, and therefore extensive destruction of thyroid tissue with a consequent decrease in risk of thyroid cancer may have occurred in the youngerindividuals (NCRP, 1985). One inference is extractable from the Marshallese data by examining the ratios of benign to malignant nodules (Table 3, column B). The Utirik population, which received no thyroxine suppression, had ratios of about 3.5:1 in children less than age ten and 3.7:1 in older children and adults, respectively. In the exposed Utirik group over the age of 10 the numberof carcinomas which developed was 3 and the number of benign nodules was 11. On the other hand, in the Rongelap group over the age of 10, which was receiving thyroxine suppression, the number of carcinomasthat developed was 4 and the numberof benign nodules only 5 with a ratio of 1.25:1. Based on the numberofthyroid cancers in the Utirik group, the number of benign Rongelap nodules in those exposed when they were over 10 years of age should have been about 15. That the relatively low number of benign nodules in this group was not the result of the higher radiation dose is seen in the plethora of benign nodules and the highest ratio of benign to malignant nodules in Rongelap children underten years ofage, all of whom received over 2000 cGy. Thyroxine suppression mayhaveresultedin the development of fewer benign nodules in the older population. (See above for the discussion concerning the limitations of the validity of this interpretation). It is not possible to determineif thyroxine prevented the development of benign nodules in Rongelap children under 10 years of age, in part because 15 of the total of 18 adenomatous nodules in this group had been detected within five years of starting suppression therapy and therefore were unlikely to have been much affected by prophylaxis. The incidence of thyroid cancer in Rongelap persons over 10 years of age was 7% and in the comparable Utirik population persons 4%, The incidence in the former might have even been higher without thyroxine suppression but this will never be proven. 9) Did fetal radiation exposure produce throid nodules? It is known that }941 given in pregnancy can produce hypothyroidism in the fetus (Fisher al., 1963). Since the fetal thyroid begins to conce iodine at about the twelfth week of pregnan of fetal thyroid injury from radioiodines begi this time. Those fetuses at the time of expos BRAVOfallout received both a whole-body d gamma radiation equal to their mothers’ a radioiodine dose to the thyroid which wasa fu of age of gestation, maternal radioiodine dose the extent of placental transfer of the radioiso The placenta is not a barrier to iodine tr (Fisher, 1975). Twelve persons followed by the Marshall Is Medical Program were in utero at the ti exposure, four from Rongelap and eight Utirik. Three of these have now developed th nodules: two of the four Rongelap children one of the seven Utirik children (the eighth p in the latter group has never been availabl examination). Table 5 summarizes the in exposure data. since the thyroid doses were not calculated t very high in two of the three (Nos. 3 and 8, 5). None of those irradiated in utero have bec@me spontaneously hypothyroid. Since the extefnal whole-body dose estimates are probably ffirly accurate, it may be that the internalized dose Iwas higher in the three persons with nodules than was estimated. All the external and internal thygoid doses calculated by Lessard et al. (1985) Were derived from a variety of data on radiation soufces and conditions, and that report states that [the maximum thyroid-absorbed dose could have be much as four times the mean values used he developed cancers that appearedafter the age of 14 _ years. One of these was a thyroid cancer wifich occurred in a person whosegestational age wag 22 weeks at exposure (Yoshimoto etal., 1988). 24