Dungy/Morgan/Heotis/ Branson/ Adams 96 turies, although interisland travel by large outrigger sailing vessels permitted the evolution of a relatively homogeneous society. The Marshall Islands were made a United Nations Trust Territory under United States trusteeship following World WarIT. In 1954, during a nuclear weaponstest series, the populations of Rongelap and Utirik atolls, numbering 253 including those in utero, were exposed to radioactive fallout from a United States thermonuclear test on Bikini atoll [4]. Since that time these exposed populations havereceived annual medical examinations and treatment carried out by the Medical Department of Brookhaven National laboratory under contract to the Atomic Energy Commission (now the US Departmentof Energy)[5]. Over the years, the Brookhaven medical team has, by virtue of its periodic field visits to the exposed populations, come into contact with a muchlarger general population which is resident in these islands. To the extent that its schedule and resources permit, the team has volunteered assistance to the local health authorities and a measure of direct patient care to the larger group. Approximately 1,500 Marshallese have been seen in this way, including a large number of children. Theyare the subject of this report. The annual medical examinations of exposed and unexposed adults have included evaluation for anemia. Hb electrophoresis has been performed on blood specimens from approximately 200 Marshallese. No variant Hbs have been detected. Splenomegaly is distinctly rare and when present has been associated with nonhematologic disorders. Malaria is not present on the atolls. Evaluation of selected individuals with microcytosis has yet to lead to a confirmed diagnosis of either a or f thalassemia. Glucose-6-phosphate dehydrogenase deficiency has not been identified among several hundred personstested. It therefore appears probable, from the results of routine evaluations of the adult population, that genetically determined anemias have, at best, a low prevalence among the Marshallese. tion-exposed parents. In the aggregate the study population repre- sented nearly 3.4%ofall Marshallese under 16 years of age. Parents or guardiansofall children were awareofthe testing and consented to their child’s participation. Copies of blood test results were placed on individual health records and givento the parent/guardian. Venous blood was collected in EDTA-anticoagulated vacutainers (Becton Dickinson, No. 6453) and skin puncture specimens were collected in 80-yl finger-stick diluting pipettes and diluted in Haema-Line 2 (Baker Instruments, Allentown, Pa.) before analysis. Hb and MCYvalues were determined by direct measurement on a Baker 500 electronic cell sizer and counter; calibration of the cell counter was donewith control cells supplied by the manufacturer. Free erythrocyte protoporphyrin (FEP) determinations were performed on light-protected EDTA-anticoagulated specimens after they had been transported at 4°C to the Hematopathology Laboratory, University of California, Irvine Medical Center. Zinc proto- porphyrin levels were assayed with an internally calibrated (hematocrit 35%) commercial hematofluorometer employing front surface illumination and digital computer output (Environmental Science Associates, Bedford, Mass.). One hundred ul of whole blood were placed on a glass coverslip and introduced into the instrument. All determinations were performedin triplicate. The laboratory’s normal range (100 subjects) is 13.8 + 10.9 ug/di and values exceeding 60 yg/dl are considered definitely abnormal. Five whoie blood controls were drawn to indicate possible changes associated with storage. The coefficient of variation for these controls was 0.3%. Statistical analysis was performed utilizing BMDP software [6] on a DEC VAX-11 computer. Among the types of analyses performed were simple and detailed data descriptions including analysis of variance and covariance,frequencies and correlations. All Hb and MCV values were used in analysis for differences among groups(i.e., island of residence, history of parental radiation exposure). Because of insufficient numbers, values of persons <1 and > 13 years of age were excluded in evaluation of normal ranges. To derive approximate age-specific normal ranges for Hb and MCV for Marshallese children the method of Dallman and Siimes [7] was used. This procedure derives limits by excluding outlying values prior to analysis. The exclusion limit for Hb ts an MCV > 5% from the mean: for MCVit is an Hb > 1 SD from the mean. ft is acknowledged that a considerable number of normal values will be amongthe excluded, but the strength of the method is that most common abnormalities will be removed. In foilowing this approachin analyzing the Marshallese data, the mean rather than median values were used because the small numberof determinations in each age group madepercentile analysesstatistically weak. The appropriateness of excluding any Hb greater than the mean was questionedin view of the rarity of pathologically high Hb levels in children. For this reason exclusion of Hb values on the basis of MCV values > 5%from the mean was recalculated to encompass only the Hblevels fess than the mean. Subjects and Methods Venousorfinger-stick blood specimens were obtained from 563 Marshallese children ranging from 6 months to 15 years of age as part of their medical examination. The recorded age wasthatof the most recent birthday. The study population resided on 4 atolls; Rongelap (n = 110), Utrik (n = 158), Kwajalein (n = 169) and Ma- juro (n = 126). Two hundred and thirteen were children of radia- IU Zab b Results Age-specific mean values for Hb and MCVin the 1- to 13-year age groups (table I) show the previously reported increase with age [8, 9]. The FEP, in contrast,