17 Table 5 Summaryof Physical Findings tn Children —Exposed (26)* Chron mpetlge active) \folluscum contagiosum Pinea versicolor Exposed un utero (4) >6 vears(30) Nonexposed Majuro(12) 0 , > 5 » Q 0 0 0 0 0 0 0 l 6 5 , 1 ; . 3 ! 0 i 0 0 3 iv -er 3cm) Palpable liver (over \denopathy 3 6 C hetlosis / Warts 5 Furuncie ash <6 years(38) 0 ; Youre OULLIS inedia S aligo Nonexposed Rongelap ° [ynea Cruces Chront ous media Nonex posed of exposed parents(24) 2 0 0 ; 5 ° 0 , ( ) , 0 ) I t 0 ; 3 °0 0 () 0 ~ 2 0 0 *Number examined. For the Rongelap population a table of most probable birth dates was eventually worked out. Although a few inconsistencies and uncertainties still persisted. these dates of birth were considered to be best estimates and were used in calculating the ages of the children for the analvses. Biologic compatibility of the birth dates within each fami- ly group was carefully checked, and physiologic compatibility of status and age for each child was examined. With the establishment of a presumptive date of birth for each child, analysis of the growth and development data was undertaken. Anthropometric data obtained during 1958, 1959, 1960, and 1961 were used in the initial analysis.* Growth data from examinations prior to 1958 had been collected by several different observers, and this earlier material will be tabulated and analyzed in a subsequent study. Although a numberofphysical and physiological parameters were measured, the present analysis was limited to stature, weight, and skeletal age. In the very young age groups head circumference data werealso evaluated. The study population was divided into 5 groups: (1) children born before the fallout and living on Rongelap at the time offallout (exposed group), (2) children born before the fallout but not living on Rongelap at the timeoffallout (control group), “The present pediatrician (W.W-S.) actively Participated in each of these surveys except the one in 1960. (3) children born to mothers who were pregnant when exposedto fallout (exposed in utere group), (4) children born subsequent to 1 January 1955 to parents one or both of whom were exposed to fallout (exposed parents group), (3) chiidren born subsequent to | January 1955 to parents neither of whom were exposed to fallout (control group for exposed parents group). Because someof the distributions encountered in these data did not grossly approximate normality or even symmetryofdistribution, and because many of the groups were too small to justify making any assumptions about the parametersof the populations from which the samples were drawn (andin many instances too small to permit calculations of any meaningful measureofvariability), all analvsis of data was done by nonparametric statistical methods.* All measures of central tendency men- tioned were medians, and all graphic presentations comparing groups wereplotted in terms of medians of the groups. Any descriptive differences between groups mentioned weredifferences between medi- ans. All tests for significance of differences between groups, unless otherwise specified, utilized the Kruskal-Wallis one-way analysis of variance.® Because the comparisonsof skeletal ages and chronological ages involved related distributions, the Walsh test'® and the Wilcoxon matched-pairs *Weare grateful to Dr. Kenneth Griffith of the M.D. Anderson Hospital, Houston, Texas, for carrying outthestatistical analysis.