TABLE 4: Distribution of thyroid nodule type by gender. Male (%) Female (%) ‘Total Adenomatous nodules 8 (25) 24 (75) 32 Adenomas 2 (29) 5 (71) 7 Occult papillary carcinomas 2 (29) 5 (71) 7 Carcinomas 1 (10) 9 (90) 10 Total 13 (23) 43 (77) 56* * The total number of nodules exceeds the number of surgeries because four patients had two categories of nodules. children one to six years of age. Three of these children were on Utirik (ages: 1, 5, and 6 years). The correlation between dose and time to development of adenomatous nodules was highly significant (r = -0.848; p < 0.001). Although the number of observationsis small, these data suggest that the earlier development of adenomatous nodules was primarily a function of higher radiation dose to the thyroid, not age. One variable which is not controlled for is thyroxine prophylaxis for the Rongelap children. Prophylaxis was notinitiated 3) What was the relation of radiation dose to time of nodule detection? The strong correlation between higher dose and earlier nodule developmentis shown in Fig.5. While this graph gives an overview of the epidemic in relation to dose, it offers little understanding of the role of the variables that shaped it. The predominant nodule type was the adenomatous nodule, the ratio of these to all other types being 4:3. Adenomatous nodules are not neoplastic. Therefore, Fig. 5 predominantly describes the relation of radiation dose to nonneoplastic nodular disease. Secondly, the prominent association of higher radiation dose with early nodule development is influenced by age-related variability in susceptibility to thyroid cancer (NRC BEIR V, 1990) and benign tumors (Ron et al., 1989; Shore et al., 1985). The mean age of the Rongelap people at the time of exposure was 27.6 years, but the range of ages was 0 (there were 4 persoms in utero) to >80 years, and susceptibility would have varied accordingly. Inferences concerning dose and time to development of adenomatous nodules can be extracted from data on persons exposed at equivalent ages, thereby controlling for susceptibility. In Figure 6 the time to developmentof nodules is graphed against dose in until ten years after exposure and after the first nodules had been detected. Therefore, the effect of thyroid suppression on development of thyroid nodules was not of consideration for the first 10 years after exposure. In addition almost all adenomatous nodules in this group had been identified within five years of initiation of thyroxine suppression. A similar age-controlled analysis for the other three nodule types is not useful because there are too few observations per group. 4) What was the relation between nodule development and age at exposure? The relation of nodule type to dose and age at exposure is shown in Fig. 7a-d. The graphs indicate a similarity in the age- and dose-related developmentof all four nodule types in the Utirik 20