196 Health Physics drink water on Kwayalem because of hus fear of the fallout, the metallic taste of water on Rongelap Island following the detonation, and the sickness (in particular, vomitmg) that some experienced after their exposure It 1s not clear how much water the Rongelap people might have been consummg on daily basis near to the trme of the urme sampling (day 16 and 17 for the Rongelap group, and day 18 for the Ailmginae group) It seems to be a reasonable assumption, however, that they were consumimg at least the mimmum amount of water necessary for healthy sustenance m tropical climates Assumingthat each adult was consummg more than 1 Ld‘ ass typical m tropical locations, one explanation for the small average urme volumes1s that the Marshallese did not provide complete 24-h urine samples to Haris However, Harms has strongly disputed this (personal communication) partly based on the assurances of Jabwe that complete 24-h urine samples were collected If mcomplete 24-h urine collections were provided, a Ingh degree of randomness between the mean and shapes of the distributions of volumes from the eight different samplings would be expected However, the distributions of urme volume were very similar in shape and central tendency, which suggests, by arguments of reproducibility, a degree of validity of the collected data Our mterpretation 1s that the volumes of urme that were sampled from adult Marshallese community members reasonably represent the true volumes excreted durmg 24-h periods of time Estimating iodine loss via perspiration. Our premise that daily hquid mtake was accompamed by sigmficant perspiration and insensible water losses resolves the apparent discrepancy between a mean value of about 500 mLcollected m urme samples and an average daily consumption of water of about 2,000 mL It 1s well established that evaporative fluid loss im the tropics 1s much higher than the values usually accepted for temperate countries (Elebute 1973) and water loss through persprration m tropical clmates can equal or exceed that lost through urme (Latham 1997) Loss of stable 10dine (normally obtamed through dietary mtake) via perspiration has been recogmzed as a sigmficant loss mechanism, 1m particularfor athletes and those living 1m hot ortropical chmates where persptration losses of body water can be large Studies im Taiwan (Mao et al 1990, 2001) of iodme Joss m sweat from athletes mdicate that 1odme concentrations m body sweat are the same before and after strenuous exercise Mao et al suggested that the lack of sigmficant differences m these various situations suggests a phystologically-based consistency to the amount of 1odme lost in sweat per umt volume In study of 13 athletes during 8 consecutive days, Mao et al found August 2010, Volume 99, Number 2 that 37 + 66 yg 10dme per L of sweat was lost and reported that average adult excretes approximately 400 to 600 mL of sweat daily through perspiration and excretes about 22 yxg iodme im the sweat The data of Mao et al (1990) represent the best known information on iodine loss (on a concentration basis) through perspiration Daily stable iodine intake.It 1s also of rmportance to make a reliable estimate of the average daily mtake of stable 1odme among Marshallese in orderto partition the datly excretion of 1odime among urime, perspiration, and feces Lattle hustorical data are available on dietary 1odme imtakes among Marshallese While iodine mtakes can, m theory, be estrmated through an understanding of diet and 1odme concentrations m foods consumed, it has been difficult for researchers to reconstruct anything but a semi-quantitative typical diet for Marshallese during the years of nuclear testmg Though some studies of foods and food mtakes have been conducted to attempt to reconstruct typical diets, those attempts have been heavily criticized (NAP 1994) for lackmg quantitative validity and for sources of possible bias To better understand typical daily 1odme mtakes among Marshallese (at least contemporarily), 1odme concentration measurements have been made imrecent years on fish commonly caught and consumed m the Marshall Islands (Takahaslu et al 1999, 2001) Concentrations of todme im fresh samples were about 100 ng g~' 1m yellow-fin tuna (Neothunus macropterus), 700 ng sg! m “reef fish,” 5,000 ng g~! 1m grant clam (Tridacnidae), and 6,800 ng g~' im mixed-type salted and dried fish Those data are reasonably consistent with other reported 1odme measurements for marme fish For example, the Chilean Iodine Education Bureau (1952), Wenlock et al (1982), Varo et al (1982), Penmmgton et al (1995), and Haldimann et al (2005) reported mean fresh weight 1odme concentrations m umdentified marme fish to be 832 (163-3,180), 750 (320-1,440), 460, 1,160 (+880), 486 (89-1,593) ng g7', respectively That mariefish have average 1odme concentrations m ther flesh (fresh weight) of a few tens of ug g”' 1s consistent with a equilibrium between their flesh and seawater which typically has an 1odme concentration of 58 pg L~' (Fuge and Johnson 1986) For Marshallese consumimg traditional diets, main- taming an adequate mtake of 1odme could only be achieved by eating marie foods smce no other foods in their diets were sigmficant sources of 1odime (Takahashi et al 2001) The average daily take of 1odime was a function of the frequency of consumption of fish, the species offish consumed, and the method of preparation For example, drymg and salting fish has been shown to imcrease 10dme concentrations about 10-fold, wile one