Normal Hematologic Values 97 Table I. Age-specific hematologic parameters in Marshatlese children Age years Hb MCV FEP MeantSD g/dl no MeantSD fl na MeantSD mg/dl no I 2 3 4 5 6 11.5+0.7 12.1+0.9 12.3 +0.9 12.341.1 12.3+0.9 12.6+0.6 29 62 46 49 57 38 77+7 7844 80+5 8045 8245 8345 29 62 45 49 56 38 30411 30 + 24 17412 22 + 23 18421 19+ 10 9 25 27 33 50 32 7 12.740.7 42 8445 42 10+ 8 39 8 9 10 11 12 13 12.84 0.9 12.9+09 12.9+0.7 13.0+0.9 13.3+0.6 13.2+0.9 35 35 35 29 34 27 84+4 esit4 8645 8644 8645 85+5 35 35 38 29 34 27 14418 10+ 11 1041] 6+ 8 9+ 9 9+ 10 28 30 3] 24 32 24 Table III. Prevalence (%) of anemia in Marshallese children, as defined by the World Health Organization (WHO)? and the 2nd National Health and Nutrition Examination Survey (NHANES 11° Age, years n tested WHO NHANESII 1-2 3-5 6-8 9-11 12-14 91 152 116 99 14.3 5.3 13.8 9.1 8.8 5.7 2.6 3.0 Males 44 2.3 2.3 (2.9) Femates 36 2.8 2.8 (3.6) 538 8.9 4.5 Total (5.7 (3.5) (2.3) (2.8) * The WHO [10] recommendeda lowerlimit for a normal Hb in children aged 6 months to 6 years of 11.0 g/dl and for ages 6-14 years of 12.0 g/dl. > Based on 95% rangesafter exclusion of subjects with abnormal! FEP, MCV or tranferrin saturation, the lower limits of normal Hb n = Number of children tested. Table II. Comparison of Hb and MCVvalues (mean + SD) obtained by finger puncture or venipuncture techniques in 2- and 3-year-old Marshallese children Finger puncture Venipuncture n tested Hb,g/dl MCV,fl 70 35 12.1+1.0 12.2+0.7 78.2447 78.8 + 4.2 was higherin the younger groups. By one-way analysis of variance there was no significant difference in age-specific Hb and MCV levels when analyzed by history of parental radiation. exposure (p = 0.24, F = 1.40 for Hb; p = 0.32, F = 1.19 for MCV). Distinct differences for Hb and MCVweredetected, however, when age-specific values were evaluated according to atoll of residence. The probability of significant difference amongatolls was p< 0.001, F = 8.77 for Hb, and p< 0.002, F = 3.52 for MCV. Thehighest levels for Hb and the lowest values for MCV were found in children living on Kwajalein atoll. To determine if significant systematic error might have been introduced by using the finger puncture technique in young children, the Hb and MCV of2and 3-year-olds who had either venipuncture or finger puncture were compared. The mean values weresimilar (table IT). An attempt to exclude variant hemoglobins, especially Hb E, was only partially suc- 9012941 Peed determined from NHANESII [11] are: <10.7 g/dl for ages 6 months to 2 years; < 10.9 g/dl for ages 3-5 years; < 11.0 g/dl for ages 6-8 years; <11.4g/dl for ages 9-I1 years, and <12.0 and < 11.5 g/dl for males and females, respectively, of ages 12-14 years. ‘ The percentages in parentheses are the prevalences of anemia found in the US by NHANESII [11]. cessful. Of 25 specimens obtained from children with lower MCV values, only 6 were satisfactory for Hb electrophoresis. All 6, with MCV values ranging from 73-78 fl, were Hb AA. The age-specific mean and median values for the Hb and MCV were found to be in close agreement. They are displayed, along with age-specific mean values after exclusion of the outlying values as described by Dallman and Siimes [7], in figures 1 and 2. A fourth age-specific curve graphedin figure 1 is higher for Hb at all points than the Hb curve derived by the exclusion limits as defined by Dallman and Siimes[7]. The difference, while small at each point (mean difference = 0.12+ 0.06 g/dl), is statistically significant by sign test analysis when all points are analyzed together (p< 0.001). Using the same approach for the MCY(i.e., no exclusion of an MCV in a child with a Hb higher than mean) no such difference was detected (fig. 2). In determiningthe prevalence of anemia two lower limits were compared; those recommended by the World Health Organization [10] and those derived by Dallmanet al. [11] (table III). The overall prevalences