Statistical observations on acute ‘radiation’ sickness in Hiroshima and Nagasaki

Before delving into the data themselves, we will note some assumptions which concern physical conditions and methods, and which will guide the interpretation of the medical data.

The seriousness of acute radiation sickness depends, above all, on the dose of the radiation received. Other important considerations are whether that dose is delivered all at once or in multiple sessions, and whether it is applied to the whole body or only to some part of it. In a nuclear detonation, irradiation should usually affect the whole body evenly, and accordingly all doses stated in the following should be taken as whole-body doses.98 Also important are type and particle energy of the radiation; this is discussed in Section 2.9.2.

The sensitivity to radiation differs greatly between tissues and cell types in the body, and therefore different organs will respond at different threshold doses. Three sub-syndromes that concern different target organs can be distinguished.

The tenet that, in Hiroshima and Nagasaki, doses sufficient to cause acute radiation sickness could have been inflicted only during the blast itself (see Section 8.1.1) gives rise to a number of testable predictions, which we will examine in the following.

8.3.1

Radiation dose as a function of distance from the hypocenter

Since there were no instruments in place to measure the radiation doses when the detonations occurred, we have to make do with approximations based on indirect methods and calculations. The officially endorsed dose estimates have seen some fairly considerable changes over time. Figure 8.1 depicts the biologically effective or equivalent doses for both Hiroshima and Nagasaki, based on current estimates of γ-ray and neutron intensities [30]. In this graph, the biological dose was calculated by applying an experimentally determined dose-dependent relative biological efficiency (RBE) function for neutron radiation [48] to the neutron component of Cullings’ radiation levels.

Figure 8.1: Estimated radiation doses at Hiroshima and Nagasaki, as a function of distance from the hypocenter. The in-air kerma for γ-rays and neutrons was taken from Cullings et al. [30]. To calculate the total dose, the dose-dependent relative biological effect of neutrons was estimated according to Sasaki et al. [48] (see text for details).

Figure 8.2: Location of survivors of the Hiroshima bombing by shielding and distance from the hypocenter. A survey in 1957 [34] canvassed all persons then living within 7 km of the hypocenter. The number of respondents who reported having been within a given distance on the day of the bombing is normalized in this graph to the size of the circular area in question. The low density of respondents who had been within 500 m of the hypocenter most likely reflects low survival rates. In contrast, the decreasing trend beyond a distance of 2 km may simply be due to lower population density in the suburbs. Data from Tables 1-4 in [34].

We will now compare observed occurrences of ARS and of survival to predicted ones. The two key sources for this purpose are Oughterson et al. [33] and Sutou [34]. Both studies report statistics on several thousand individuals. The first one was compiled by the ‘Joint Commission for the Investigation of the Effects of the Atomic Bomb in Japan’, a group of American and Japanese physicians convened at the initiative of Ashley W. Oughterson, a professor of surgery at Yale who at the time was serving as a colonel in the U.S. military. This commission only arrived at Hiroshima and Nagasaki in October 1945, but it did acquire and organize data previously collected by Japanese physicians; and the statistical evaluation of these earlier Japanese data constitutes the main substance of the commission’s report. Most of the figures tabulated in [33] pertain to patients still alive and in medical care at 20 days after the bombings; recorded are slightly below 7000 survivors in each of the two cities.100

The second study was carried out in 1957 by Dr. Gensaku Oho,101 a physician from Hiroshima, who enlisted the help of student volunteers to canvas the resident population of Hiroshima. The main purpose of this study was to determine the occurrence of radiation sickness among persons who had not been exposed to the bombings themselves, but who had entered the area close to the hypocenter only afterwards. The more recent paper by Sutou [34], which is used here, is a partial translation of and commentary on Oho’s earlier study.

The observations made above for Hiroshima mostly apply to Nagasaki as well (see Table 68N in [33]); however, some findings are quantitatively more pronounced. Symptoms of and death due to ARS are less frequent within 1 km than in Hiroshima, even though radiation doses are supposed to have been higher (see Figure 8.1): among survivors exposed in the open or shielded only by a wooden house, less than 60% exhibit epilation or purpura. Among survivors exposed between 1.5-2.5 km, a greater percentage than at Hiroshima shows symptoms of ARS. On the other hand, beyond 4 km from the hypocenter, that percentage does indeed drop to zero in Nagasaki, whereas it remains positive even at this distance in Hiroshima.

Concrete buildings will afford substantial protection from both γ-rays and neutron radiation, and we should therefore expect a lower number of ARS victims among those inside these buildings than in those inside wooden buildings or in the open. This is indeed observed; within 1 km from the hypocenter, the incidence of ARS is approximately 25% lower inside heavy buildings than outside, both in Hiroshima and Nagasaki (Oughterson et al. [33], Tables 68H and 68N). Yet, ARS inside heavy buildings in Hiroshima remains more prevalent than it is in the open in Nagasaki, even though the radiation dose is said to have been higher in Nagasaki.

More detailed statistics on this question are reported by Oughterson and Warren [146], who in their Table 3.7 show findings from three individual concrete buildings in Hiroshima, all of which were situated between 700 and 900 meters from the hypocenter. In each building, some people were protected by multiple walls or floors, such that the total shielding was equivalent to ≥ 154 inches (or 394 cm) of water (see Table 8.2). The stated radiation dose outside the buildings was up to 80 Gy, which amounts to approximately ten times the lethal dose. However, after passing through this much shielding, it should have been attenuated to a mere 4 mGy. This corresponds to just 2/3 of the typical annual dose of a U.S. citizen and will, of course, not produce any acute symptoms at all.

Nevertheless, Oughterson and Warren report cases of ARS—some of them lethal—among persons thus protected. They propose that these may be due to neutrons, apparently assuming that neutrons are less effectively shielded by concrete than are γ-rays. However, this is now known to be incorrect (see Table 8.2); and moreover, as already noted, the estimated neutron dose at Hiroshima was very substantially reduced in the decades after their book was published [49].

As a second deus ex machina, the authors suggest that the bomb’s γ-radiation may have been of much higher particle energy, and therefore more penetrating, than is generally assumed. However, they do not offer a physical basis for this hypothesis, nor do they pursue its wider implications for the physical and medical dosimetry of the entire event, which would have been substantial. Such lack of thoroughness suggests that the authors themselves do not take their own proposal seriously. When commenting on the reverse scenario—the wondrous survival of some individuals exposed to strong γ-radiation—the authors dispense with any special pleading and blankly state (p. 63):

It is equally difficult to explain the complete absence of radiation effects in a number of people who were theoretically exposed to lethal dosages of radiation.

We note that Oughterson and Warren acknowledge the dilemma of ARS occurring among those beyond the reach of the bomb’s radiation, while failing to appear in some of those exposed to a ‘theoretically lethal’ dose. Adjusting dose estimates will not solve this dilemma: increasing doses may avoid the Scylla of death despite protection, but it will wreck the ship on the Charybdis of inexplicable survival; assuming lower doses to explain miraculous survival will make the deaths of shielded victims all the more incomprehensible.

The occurrence of ARS symptoms in persons who were outside Hiroshima on the day of the bombing, but who entered the zone within 1 km of the hypocenter afterwards, is a crucial piece of evidence. While anecdotal reports are found in many sources [12,14,16,32,62], the only statistical survey on this question is the one by Gensaku Oho; and it is telling that we owe this crucial study to the personal initiative of this energetic doctor from Hiroshima and his student volunteers, rather than to the official institutions created and maintained for such investigations by the governments of the United States and of Japan.

Figure 8.3: Symptoms of ARS in 525 persons who were outside Hiroshima during the bombing, but who came within 1 km of the hypocenter in the aftermath, as a function of time spent in that area. Data from Table 7 in [34]. The regression line was fitted with weighting for sample size. Symptoms of ARS include fever, diarrhea, bloody stools, bleeding from the mucous membranes, loss of hair, and generalized weakness.

Oho’s most important findings are summarized in Figure 8.3. Many people entering the area within 1 km of the hypocenter report symptoms of ARS; the percentage of people thus affected exceeds 50% among those who stayed for more than 2 days. Additional tables and figures presented by Sutou [34] clearly document that the same effect is also present among those who were in Hiroshima during the bombing: while of course many in this group suffered ARS regardless of their whereabouts in the aftermath, the incidence is higher among those who also came near the hypocenter in that period.103

Findings such as those reported by Oho can, of course, not be explained with the radiation released during the detonation. There are three ways of dealing with this problem:

1. The findings are ascribed to fallout or residual radiation, which are assumed to have been much greater than conventional or official estimates [34,158,159].
2. The findings are declared to ‘warrant further analysis’ and then studiously ignored [36, p. 90].
3. The findings are ignored without ceremony. If you guessed that this is the most common approach, you are indeed correct.

The last two alternatives require no further comment. Regarding the first one, it was shown earlier that real fallout must have been lower, not higher than the official estimates, and there is no basis whatsoever for higher estimates of neutron-induced radioactivity.104

The thesis of this book—namely, that sulfur mustard, not radiation was the cause of ‘ARS’—provides a ready explanation for cases of the disease among late entrants to the city. Sulfur mustard is known to linger, and its persistent stench was noted by Burchett four weeks after the bombing [16]. Wind-driven mustard fumes would explain why those located downwind from the hypocenter suffered more ARS [158] and were at greater risk of developing cancer [160,161]. While Yamada and Jones [158] ascribe the surplus incidence of ARS in this group to high β-radiation from isotopes contained in the black rain, the very low levels of ¹³⁷Cs in extant black rain samples [6] clearly disprove their explanation.105

In patients who suffered ARS due to exposure only after the bombing, the symptoms should develop with some delay; and this is indeed reflected in the statistics reported by Oughterson et al. [33].

Figure 8.4: Time of onset of purpura and oropharyngeal lesions in Hiroshima bombing victims, and blood cell counts in accidentally irradiated patients. Data for onset of purpura (bleeding) and oropharyngeal lesions in Hiroshima victims from Table 17H in Oughterson et al. [33]; 100% is the total of all patients that exhibited the symptom at any time during the observation period. Platelet and granulocyte counts (from Fliedner et al. [164]) represent median values of 11 patients who were exposed to whole body irradiation at Chernobyl. Values are relative to those on day 1, which were in the normal range for both cell types.

Characteristic symptoms of ARS hematopoetic syndrome (see Section 8.2.1) are purpura, caused by the failure of the blood platelets, and oropharyngeal ulcers due to bacterial and fungal infections, which are brought on by the lack of granulocytes. In patients who exhibit these symptoms after exposure to a single dose of irradiation, they become manifest between days 8 and 28, with shorter latency at higher doses [149]. Figure 8.4 shows that this is also true of most Hiroshima bombing victims; however, in about 25%, the initial manifestation is delayed until the fifth week or later.106 For illustration, the figure also shows the time course of platelet and granulocyte counts in patients exposed to irradiation after the reactor meltdown at Chernobyl. Both cell counts reach their lowest point before the 28^th day, which explains that symptoms will be manifest by this time.

Anecdotal evidence confirms the occurrence of late cases. For example, in his posthumously107 published book First into Nagasaki [166], the American journalist George Weller notes on September 22^nd:

New cases of atomic bomb poisoning with an approximate fifty percent death rate are still appearing at Nagasaki’s hospital six weeks after the blow fell … Whereas formerly twenty patients a day with dwindling hair and their bone marrow affected were coming to Japanese hospitals, the rate is now fallen to about ten.

The decreasing, yet still ongoing observation of new cases agrees with the data in Figure 8.4. While from this limited information we cannot be sure whether the death rate in new cases was indeed falling, this would be plausible in real ARS [149] and similarly also in mustard gas poisoning. What is not plausible in true ARS, however, is the repeated occurrence of new cases, particularly ones with fatal outcome, as late as six weeks after the exposure. These patients must have taken in the poison some time after the bombings, probably in a cumulative fashion, just like some of the subjects surveyed by Oho [34].108

You may have seen studies on atomic bomb survivors that correlate some biological outcome such as cancer with individual radiation doses. The question of dosimetry will be discussed in Chapter 11, which also shows a graph which correlates the incidence of ARS symptoms radiation doses (Figure 11.1B). The correlation is obviously very poor, and the dose-response curve is wildly implausible biologically, as can be seen by comparison with proper data shown in Figure 11.1A.

Figure 8.5: Numbers of survivors grouped by dose values (A), and incidence of ARS symptoms among those assigned an estimated dose of 6 Gy (B). In RERF’s dataset [168], one or more symptoms were given as ‘not reported’ for 8 out of the 72 survivors with exactly 6.000 Gy; these subjects are included in A but excluded in B.

The data set from which Figure 11.1B was constructed contains radiation doses at higher resolution than depicted in the figure. If we plot a histogram of the number of people grouped by the individual dose values in the file, we see that the dataset contains no cases with estimated doses above 6 Gy (Figure 8.5 A). However, the number of people with an assigned dose of exactly 6.000 Gy greatly exceeds that of any other individual dose value above 3 Gy; in fact, only below 1 Gy do we find dose values with higher head counts than 6 Gy exactly. This peculiar pattern strongly suggests that all raw dose estimates higher than 6 Gy were simply truncated at that value; probably because they were deemed unsurvivable, and quite possibly under the impression of the rhesus monkey experiments shown in Figure 11.1A. It should go without saying that such sausage-making does not qualify as science. Furthermore, whether truncated or not, in this highest of all dose groups, the number of individuals with 0 or only one symptom of ARS exceeds that with two or more symptoms (Figure 8.5B). The 22 individuals without any symptoms clearly count among Warren’s and Oughterson’s mystery patients with ‘complete absence of radiation effects’ in spite of exposure to ‘theoretically lethal’ doses of radiation.

The findings presented in this section reinforce our previous observation that the distribution of ARS does not fit the official story of the bomb and its radiation. We will revisit the question of purported radiation doses and biological effects in Chapters 11 and 12.

Before leaving this topic, one recurrent motif in the reports on ‘radiation sickness’ from Hiroshima and Nagasaki should be noted: the widespread and early occurrence of diarrhea, often bloody, among the patients. A graphic account is given by Michihiko Hachiya [62]. The author, a head physician who had been injured in the bombing and admitted as a patient to his own hospital, wrote in his diary on August 7^th:

Everything was in disorder. And to make matters worse was the vomiting and diarrhea. Patients who could not walk urinated and defecated where they lay. Those who could walk would feel their way to the exits and relieve themselves there. Persons entering or leaving the hospital could not avoid stepping in the filth, so closely was it spread. The front entrance became covered with feces overnight, and nothing could be done for there were no bed pans and, even if there had been, no one to carry them to the patients.

Disposing of the dead was a minor problem, but to clean the rooms and corridors of urine, feces, and vomitus was impossible.

Such events would suggest an outbreak of some virulent enteric pathogen, which is indeed common in disaster situations; and Hachiya and his staff initially assumed this to be the case. Also on August 7^th, Hachiya writes:

Dr. Hanaoka … brought word that there were many who not only had diarrhea but bloody stools and that some had had as many as forty to fifty stools during the previous night.109 This convinced me that we were dealing with bacillary dysentery and had no choice but to isolate those who were infected.

Dr. Koyama, as deputy director, was given the responsibility of setting up an isolation ward.

However, already on August 13^th, he notes:

My conjecture that deaths were due to the effects of a germ bomb causing dysentery I had to discard because diarrhea and bloody stools were decreasing.

Hachiya’s conclusions are confirmed by the data given in Oughterson et al. [33], which show that both bloody and non-bloody diarrhea are strongly correlated with other ARS symptoms, and also that case numbers were highest early on and then declined (Figure 8.6), even though the hygienic conditions remained about as bad as can be imagined.

Figure 8.6: Time of onset of diarrhea and vomiting in Hiroshima bombing victims still alive 20 days after the bombing. Data from Table 17H in [33]. The first data point in each series represents the day of the bombing.

Diarrhea can indeed occur in real radiation sickness. However, it commonly occurs very early on only in patients who have received a dose of 6 Sv or greater [149]. Under the conditions then prevailing in Hiroshima and Nagasaki, patients hit with such a high dose would not have survived. Yet, the data listed by Oughterson et al. [33] pertain to patients who were alive 20 days after the bombing, and 96% of whom remained alive when the study concluded several months later (see Section 8.4.2), which means that they were not lethally irradiated. Thus, the timing of diarrhea observed in Hiroshima also indicates that the reported radiation sickness was not actually caused by radiation. On the other hand, early onset diarrhea has been described in multiple reports on mustard gas exposure of humans and of experimental animals (see Section 7.3.5).

Many of the data presented in this chapter were drawn from the report of the Joint Commission [33], and we saw that these data contain clear evidence against nuclear detonations as the cause of ARS in Hiroshima and Nagasaki. We thus might wonder what the commission’s members, most of whom were physicians, were really thinking while they assembled their data. The only first-hand account by any of them which I have found is that by the pathologist Averill Liebow [77]. The author offers many interesting glimpses into the conditions of the work, but he does not betray any doubt or puzzlement concerning its scientific findings. However, writing originally in 1965, Liebow notes:

It is true that few who took part are left to tell … Indeed it is as though some curse, like that which the superstitious say fell upon Lord Carnarvon and his men when they violated the tomb of Pharaoh Tut-ankh-amen, has been visited upon those who pried into the ravaged heart of Hiroshima. Only three of the seven American medical officers live. Drs. Oughterson and Tsuzuki, the chief organizers for the two countries, have died; so too, while still young, have Drs. Calvin Koch, Jack D. Rosenbaum, and Milton R. Kramer. May this record do honor to these able and devoted men.

Liebow’s analogy surely is intriguing. We will, however, leave it for others to pursue, lest we be accused of superstition.