Memo to H Pratt, Subject: Enclose Report " a Reconstruction of Chronic Doses for Rongelap and Utirik Residents - 1954 to Present "

Page 1

BROOK HAVEN NATIONAL LABORATORY To: Hugh Pratt of data collec~ed by yoursc,’.iand member= of your medical team. Your comments would be Sreally appreciated. traveling t:> the Marshall Lslan<s,’ and to Seattle, k’ashingtm to the Annual Ysalth Physics Societ> Meeting. Thus, your timely review will afford IS the opportunity ~c incorpor~te changes in the final

Page 3

13/Q. Four persons visited Argonne from Rongelap and, in addition, pooled urtlw samples from both atolls were analyzed radiochemically for Cs and hr. Subsequent Brookhaven National Laboratory expeditions by the Medical Department and Safety and Environmental Protection Division utilized whole body counting and radiochemical analysis of urine and blood samples to identify and quantify the radionuclides that were present in the body. The results of these radiological measurements are given in terms of body burden in tables The aforementioned body burden tables illustrate adult mean values for Rol~gelap and Utirik. An adult, as classified here, was a person over 16 years of age. The mean body mass in this age interval was 60 kilograms. The observed body mass versus age distribution is shown in figure 1 for Rongelap residents. The same body mass versus age distribution was observed at Utirik. Due to the paucity of measurements at Utirik, information on 6oCo, 65Zn and Fe was in some instances derived from the ratio of adult mean body burdens between Rongelap and Utirik. A mean ratio of 2.6 was observed in body burdens for65Zn, 90Sr and 137Cs after they reached their maximum values. The standard deviation of this ratio was 15%. In the foll~ing analysis, personal body burden histories and residence intervals, in conjunction with contemporary dosimetric models, are used to’estimate internal dose. Dosirnetric distributions were constructed from the results and a suxmnary of the derived activity ingestion rates and dose equivalents was provided for various subgroups of the population. Additionally, exposure rate history curves were constructed for each atoll for the period following the BwV() test. These data, together with appropriate conversion factors and living patt:rn models, provi(led an estimate of external dose equivalent.

Page 6

Declining continuous uptake of radioactive dietary items was m;lthematicall modeled for each nuclide of concern. The following general eq~lations were t S time post onset of uptake, days, A Z instantaneous fraction of atoms decaying per unit time, day-l Ki ~ instantaneous fraction of atoms removed from compartment i by physiological Xi ~ compartment i deposition fraction, x: = the number of atoms in compartment i relative to the number in all compartments at the onset of uptake (t=O), u E instantaneous urine activity concentration, Bq 1-1, Us ~ subject urine excretion rate, i day-l,

Page 7

‘1 s traclion from GI tract to blood, f E fraction excreted by the urine pathway, KE E instantaneous fraction of atoms removed or added to the atom uptake per q“ Z body burden at the onset of uptake, Bq, D: the number of disintegrations in all compartments occurring during the uptake interval, Bq days. Equations 1 and 2 were used to determine the dietary removal rate constant K ~ and then the initial daily activity ingestion rate required to produce the measured or derived body burden. number of disintegrations that occurred in val of an individual living on Rongelap or DETERMINATION OF KE Equation 3 was used to determine the the body during the residence interIf the mean residence time in the diet is dence interval, then constant continuous uptake much much longer than the resiis achieved. Equations 1 and 2 can be converted to the constant continuous equations by replacing ~g vith -A. Single uptake expressions are obtained by setting P“ equal GO zero, In some cases only radioactive decay may remove the nuclide frcm dietary items; for these cases KE would equal zero. In the ‘case of the former Bikini ~eeidents, the maturing of coconut trees during re$idence on Bikini Atoll caused @ continuously increasing dietary uptake of . Thus, KE was found to have a negative value. In the case of Rongelap and Utirik, KE was found to have a posiradioactive decay, some other removal mechanism decreased the radioactivity in

Page 8

d Jry iL,,nlsduring the residence interval. For the nuclide 55Fe, only one m sl.]trnentillblood was published by the BNL Medical Program (Be72); thus an est ‘.ll.lt, of KE was not possible. KE was determin(:d by using equation 1 or 2 and the population subgroup n,<~ii,body burdens or urine activity concentrations. A portion of these bioass d~Lr are iIllustrated for adult males and females in Figures 2, 3, 4, 5 and 6. TWO conse ,,tive urine or body burden data points were used to eliminate the un. knoi.n ing! tion rate from the equation. This method yields n-1 estimates of K Wliere n WJ,. the number of data points. An average value of KE was assigrled fo each nuclide and the results for the Rongelap and Utirik populations are given ir-rTable 3. For the evaluation of KE from equations 1 and 2, radiological and physiological parameters were obtained from the open literature (ICRP59, ICRp6 IcRP69, ICRP79, Ki78). A representative sample of these parameters is present

Page 9

Total Body Dosimetric and Physiologic Data Capartment Cuapartmenc GI Tract Fraction Deposition Removal to Blood Excreted in Decxy Significant Branching Fraction Rate Constant Transfer Urine Conacaoe Progeny Ratio

Page 10

The values of KE were similar for males and females and for residents of Rongelap and Utirik. For 90Sr on Rongelap a factor of 2 difference between KE values was observed for males and females. The female parameter for Rongelap Atoll compares with that obtained from the Utirik data. A paired t-test of the Rongelap male and female data indicates that the male/female difference was highly probable and therefore not significant. This difference leads to a bimodal activity ingestion rate distribution for 90Sr in the Rongelap populaData for 60Co and 65Zn was not sufficient for analysis for the Utirik Atoll residents. Values for KE observed at Rongelap were assigned to Utirik males and females and body burden histories for population subgroups were reconstructed using equation 1 or 2. Figures 7 and 8 illustrate the derived mean adult body burdens for all significant nuclides studied on Rongelap and Utirik. This method provides a best fit of the data shown in figures 2 through 6, and provides a body burden history during Utirik, a time when body burden measurements are also plotted to demonstrate the fit. The curves shown for 55Fe in figures 7 the early years post return at were not made. Actual data points and 8 were obtained by setting KE equal to zero. This underestimated the initial body burdens and overestimated future ones. Since 55Fe contributed less than 1.0% to the total dose equivalent an arbitrary assignment of KE based on observed values for the other nuclides was not attempted. During 1974, another series of blood samples wae obtained from Rongelap and Utirik (C075). Analysis for 55Fe was to be performed; however, no records have been located. Once these records are found, a reanalysis of 55Fe and its impact on early dose equivalent rates will be conducted. A substantial change in dose equivalent is not to be expected.

Page 11

Figuve 4 and figure 6 illustrate the observed adult histories of Sr and 137Cs mean urine activity con(:entrations. Mean values for adult males or all adults were plotted. Measured values for 137Cs body burdens were also shown in figure 7. A much smoother curve was plotted in figure 7 and it was determined that the collection and analysis technique for urine samples introduced the additional variations. Based on this observation for 137Cs, a smooth body burden curve for Sr, reconstructed from raw data and equation 1, was considered a more accurate history. A detailed presentation of the greater variation in urine bioassay measurements versus direct body burden Figure 9 illustrates the variation exhibited in measurements can be found the body burden of 5 randomly chosen subjects over the 25 year monitoring period. These individual variations may have had a dramatic impact on the mean data. In figure 2, which illustrates the adult male, adult female and adult burden for the 25 year exposure period, a decrease seen during the years 1958 through 1963. Although population mean Cs body followed by an increase was the Castle BRAVO test initially contaminated Rongelap, it had been proprosed that the Hardtack Phase I series added to this an amount of co~tami~ation equal to that respo~aible for the figure 2 body burden pattern (c063), Figure 9 suggests that moat individuals counted in those years had maintained or declined in body burden; however, one individual’s burden (#881 M) rose and fell quite differently from the others. Several factors could have contributed to this variation from the mean such as departure and return to the atoll, sickness, the dietary contribution of imported foods, etc.. Since the mean values are based on small numbers of persons who were chosen at random, it is conceivable that individuals like 881 M

Page 12

influence(l the mean body burdens to a greater degree than recontamination of the inhabited atolls. This conclusion is moot since the body burden of an individual was not monitored consistently throughout the residence interval except in the few cases exhibited in figure 9,

Page 13

RESULTS AND DISCUSSION Daily activity ingestion rates were calculated for dosimetrically significant nucli(l(s post return. An exponential decline was proposed for the ingestion rate ~ithin a population subgroup and initial reference values are given in figures lG, 11, 12, 13 and 14 (June 1, 1957 was assigned as a return date to Rongelap). Figure 10 demonstrates the difference in ingestion of Cs for various population subgroups. This undulating pattern was exhibited by and ‘5Zn, nuclides for which sufficient data existed for analysis. Differences in ingestion rates of the stable element at the same geographic location have been shown to occur among members of a population (ICRP 23). Age dependent diet studies for ingestion of CS for urban Japan have values varying from 11 Wg d for adults to 8.6 pg d for children. Sr in a westerntype diet ruse from 600 pg d for infants to 690 Vg d for 5 year olds to 3,600 Mg d-1 for 13 year olds and fell to a mean of 1,900 Pg d-1 for adults. Zn in the United Kingdom rose from 2 to 40 mg d , the higher value of Zn being observed in adult tea drinkers. Fe ingestion in a western-type diet has a minimum at age 3 and maximums at ages 1 and 20 years. Co is ingested at a rate of 20 Pg d-l for Japanese adults and half this amount for children. The Marshallese population also exhibits dietary changes as a function of age. The authors of the Marshall Islands Diet and Living Pattern Study (Na80) observed COconut sap being used as a major food supplement for infants, and later in adult life as a major source of daily fluid intake. Since coconuts and coconut tree sap provided the major source of Cs on Bikini Atoll (Le80, Mi80), the shape of Figure 10 was in agreement with the qbserved diet pattern.

Page 14

Fig,lre 11 compiled the individual data calculated for 137CS for all Rongelap Iesidents and was referenced to June 1, 1957. The individual maximum 137Cs daily activity ingestion rate was approximately 4 times the population mean valut,. The standard deviation observed for the adult activity ingestion rate distribution was 41% of the mean value, 39% of the mean value for young adults, 48% for adolescents, 38% for children and 54% for infants. Adolescents and infants exhibited a broader distribution than adults while children showed a fractional variation in activity ingestion rate similar to adults. Breast feeding versus coconut sap supplements would have contributed to the greater varia:!.on observed in infants. Adolescents and young adults were the population subgr nips which have been observed to move frequently between atolls. This mobility would lead to greater variations in the daily activity ingestion rates rela?:ve to those observed in the more stationary population subgroups. Figure 12 also exhibited a wave pattern; however, a distinct difference between males and females was indicated. This difference arose from the use of dietary rate constants listed in Table 3 which were derived from urine data for male and female residents at Rongelap Atoll. Its major impact vas on the dose equivalent rate, not on the total dose equivalent; and its effect was to cause the dose equivalent rate for males to rise and decline more rapidly than for F!gures 13a and 136 summarize the individual data for 90Sr for all Rongelap residents and were referenced to June observed for the distributions which contained difference in the 90 Sr dietary rate co>stants. indicatej that the observed difference between 1, 1957. A bimodal shape was both sexes thus reflecting the Data from urine bioassay the male and female value for K=

Page 15

ment data Juring the twenty -tliree year residel~ce interval. The results indicate that because of urine activity concentration variability, a probability of 6 out of 10 existed such that the m.~le value for KE would be different from the female value by the factor observed. Thus differences in the derived activity ingestion rates and dose equivalents were not significant. Figure 14 shows a semi-log plot of the 65Zn and 137Cs activity ingestion rate histories for adults on Rongelap. A smooth curve was drawn between points and the appearance of an increasing Cs ingestion rate during the 1960’s indicates the possibility of anotl~er ries was conducted just prior to from the Cactus, Yellow Wood and contaminating event. The Hardtack Phase I sethe observed increase in the curve and fallout Hickory experiments detonated at Bikini and Enewetak would have reached Rongelap. However, several observations fail to support the conclusion that recontamination was significant. These are as follows: 1) the increase in 137Cs ingestion rate was not in conjunction with an increase of 65Zn; however, since 65Zn is an activation product it may have not been produced in the same proportions 2) the peak Cs body burden at Utirik occurred nearly three years after the initiating event, Castle BRAVO, while the peak body burden at Rongelap followed six years after the potentially contaminating experiments of the Hardtack series in 1958 3) the activity ingestion rate at Utirik demonstrated a continuously declining pattern versus the humped pattern observed at Rongelap. This occurred even though there was an equal external exposure rate hist~ry following the Hardtack series as measured by the U.S. Public Health Service on both Rongelap and Utirik (Un 59). These facts suggest that the Hardtack series was not a major factor influencing the Rongelap body burden patterns. Thus it is assumed that persons who had body burdens significantly differel]t from the mean body burden for the population

Page 16

caused th( , Lent of variation reported. Based on these observations a smooth descripti, II 1 the body burden and activity ingestion rate was adopted and a declining co linuous uptake model was used to generate the curves in figures 7

Page 17

INTERNAL [~~Jsii EQUIVALENT RATES The ~pproximate instantan(:ous dose equivalent rates for the Lotal body were detemined from the …