TAEA-SM-224/607 109 Basal serum TSH concentrations and response to TRH Since the most sensitive index of impaired thyroid function is an elevation in serum TSH which occurs through the hypothalamic-pituitary-thyroid feedback axis, serum TSH concentrations and their response to TRH were measured in both the control and the exposed Rongelap population. In primary hypothyroidism, the response of the pituitary to TRH is excessively great [8]. Mean basal TSH was 2 wU/ml in 25 non-exposed euthyroid Marshallese, and the range was from undetectable (<0.05 U/ml) to 3 pU/ml (Table IV). Serum TSH 20 minutes following TRH wasincreased in all control subjects. The mean increment was 11.5 + 4.5 (SD) with a range of from 4.7 to 20 wU/ml. Theseresults are not significantly different from those previously reported in other populations[9]. On thebasis of these studies, criteria were established for classification of patients as having biochemical evidence of impaired thyroid function. These criteria are summarized in Table V, and include either two basal TSH determinations greater than 5 wU/ml (> 4 standard deviations above the mean) or basal plasma TSH >3 pU/ml (but <5 wU/ml) and plasma TSH afterTRH > 22 pU/ml. Consistent observations in these ranges were required on two occasions to meet the criteria for biochemical evidence of thyroid dysfunction. While serum T, concentration is an important determinant in the thyroid status of the individual, previous studies have indicated that evidence of impaired thyroid function can be elicited by these tests before serum T, concentrations have fallen below the normal range [10]. Therefore, the serum T, concentration was not used as a criterion in establishing the diagnosis of impaired thyroid function. In Table VI is shown the frequency of at least a single elevated basal TSH concentration in various Marshallese populations. In a control group of 115 who were not exposed to radiation, 11 subjects or 10% of the population had a serum TSHgreater than 3 pU/ml. In ten of these, serum TSH was only minimally elevated (4.0 nU/mlor less); the remaining value was 6.1 wU/ml. None of these patients had detectable clinical hypothyroidism or thyroid enlargement, but serum T, concentrations were generally in the low normalrange. In the exposed Utirik population, 12 of 99 subjects tested had at least one basal serum TSH greater than 3 nU/mil, though none of these was in excess of 5 pU/ml. The incidence of elevated TSH in this populationis not significantly different from that of the unexposed group. In the Rongelap and Ailingnae population, 11 of 43 subjects were found to have at least a single elevated basal serum TSH greater than 3 «U/ml, and in two cases serum TSH was in excess of 7 U/ml, and in two cases serum TSH was in excess of 7 wU/ml. This is a significantly higher prevalence than in the other two groups pooled (p < 0.05). In Fig. | are shown the responses to TRH ofthe four individuals who met the criteria given in Table V. The normal basal TSH and response to TRH are shown in the shaded bars. In these four individuals, the basal serum TSH waselevated,