THE CRITERIA OF RADIATION EMPLOYED IN THE STUDY (2024)

THE present chapter will be devoted to an examination of the available and practical criteria for evaluating the amount of radiation received by individuals exposed in Hiroshima and Nagasaki to the effects of the atomic bombs.

4.1The complicated nature of the injuries sustained by some survivors; “disaster effect” vs. “radiation effect.”—The detonation of an atomic bomb may be harmful to persons in its vicinity for a variety of reasons. The air blast may inflict injury directly or secondarily, from flying debris and falling walls and roofs. The thermal radiation produces severe burns, and, in addition, further burns may be sustained as a result of fires kindled by the explosion. Under the conditions prevailing in Hiroshima and Nagasaki, serious secondary infection of these burns was the rule rather than the exception. Finally, the explosion is accompanied by the release of a variety of deleterious high energy radiations, particularly X-rays, gamma rays, and fast and slow neutrons. The complex manner in which these various noxious agents affect persons exposed to the explosion of an atomic bomb is well depicted in the Report of the Joint Commission for the Investigation of the Effects of the Atomic Bomb in Japan (Oughterson et al., 1951).

Moreover, in the wake of such a cataclysmic event as the detonation of an atomic bomb over a city, come certain well recognized accompaniments of any large-scale disaster, notably, disturbed nutrition and increased morbidity from disease. These “disaster effects” are here so complexly intertwined with the primary event that it will be difficult, if not in fact impossible, in any analysis of atomic bomb sequelae, to effect a realistic and completely satisfactory separation of the relative importance of what transpired at the time of the bombing and what transpired in the next several months. In this particular instance, the general problem of disentangling primary and secondary effects upon the manner in which pregnancies are terminating is still further complicated by the fact that certain delayed somatic effects of irradiation have been shown to occur in the population under study. The “late” effects include leukemia (Folley, Borges, and Yamawaki, 1952; Lange, Moloney, and Yamawaki, 1954; Moloney and Lange, 1954) and cataracts (Cogan, Martin, and Kimura, 1949; Sinsky, 1955). Although both of these conditions affect only small numbers of the more heavily irradiated survivors, they do occur in the group most critical to this study. It seems entirely possible that there are other late sequelae, still undefined, which may influence reproductive behavior.

In subsequent chapters an attempt will be made to correlate exposure history with a variety of possible indicators of genetic damage. However, it must be clearly understood that if a significant association can be shown to exist, this demonstration of itself does not enable one to conclude that exposure to the bomb has genetic consequences; the latter conclusion is justified only if an effect remains after the various other factors associated with the irradiation have been assigned their proper role.

4.2The question of residual radiation following an atomic bombing.—The ionizing radiation of an atomic bomb explosion is released over a relatively brief period, of approximately 100 seconds. However, as a result of the contamination of the area due to the fall-out of radioactive by-products of the explosion, as well as secondarily induced radioactivity, there is a certain amount of residual radioactivity in an area over which a bomb has been detonated. The estimation of the amount of residual radiation at Hiroshima and Nagasaki presents many difficulties. The distribution of residual radiation around the hypocenter in the two cities was asymmetric, the exact pattern depending on local meterological details (Warren, 1945; Warren, 1946). Based on the observed residual radioactivity at intervals following the bombing, Warren (1945) has estimated that the maximum irradiation due to residual radioactivity was in Hiroshima during the first 60 days following the bombing the equivalent of 4.2r, and in Nagasaki during the first 47 days, the equivalent of 14.2r. Wilson (1951), using similar data, has suggested that in Hiroshima the maximum cumulative dose of residual radiation at distances greater than 1,000 feet (300 meters) was not more than 10r, with somewhat higher doses prevailing in Nagasaki, especially in a fall-out area extending eastward from the hypocenter, where the cumulative dose from radioactivity might have amounted to 100r. Irradiation of this latter degree would only be experienced by individuals establishing continuous residence near ground zero immediately after the bombing, a situation which rarely, if ever, obtained, although persons coming into the area later or intermittently might of course receive lesser amounts. On the other hand, there are persistent Japanese reports of members of rescue parties and others not actually in the two cities at the time of the bombing later developing symptoms of radiation sickness, or leucopenia (Japan Science Council, 1951; Appendix 18, ABCC Semi-Annual Report, 1 January 1952–30 June 1952); this would suggest a dose of 200r (cf. Sec. 4.8). This apparent conflict of evidence remains unresolved at present; we shall adhere for the time being to the more conservative view concerning residual radioactivity. By and large, individuals exposed to the effects of the atomic bombs tended to leave the area as rapidly as possible. Accordingly, for the purposes of this study it has been felt that although exposed individuals and also those entering the city immediately after the bombings may have been subjected to some residual radiation, by comparison with the amount absorbed by those present at the moment of the explosion, this was on the average small and could be disregarded. Moreover, it might be pointed out that in the case of an individual engaged, e.g., in rescue work in Hiroshima or Nagasaki during the first few days following the bombing, it appears impossible to reconstruct his behavior during those several days in a fashion which would permit a realistic estimation of the total radiation dose which he absorbed during this period.

4.3Factors determining the nature of the radiation data collected in this study.—At the time this study was being planned, in 1947, no data of any sort concerning the types of radiation emitted by an atomic bomb and the distance-dosage relationship were available to any of those closely connected with the program. In retrospect, in view of the many revisions the physical estimates have undergone, this is perhaps fortunate. Be that as it may, under the circ*mstances extensive recourse was had to the findings of the Joint Commission, then available in preliminary form. These findings, as later published (1951), established the following facts of importance in the design of the present observations:

4.3.1The syndrome of radiation sickness due to whole-body irradiation.—Although the therapeutic uses of radiation had long since laid the groundwork for an appreciation of the signs and symptoms of “radiation sickness,” it remained for the observations of the Joint Commission to provide the definitive description of this entity. The findings, of course, vary according to the radiation dose. We are not here concerned with the early effects of lethal doses of whole-body irradiation, since persons receiving such doses can scarcely provide the parentage for a study of this nature. The Joint Commission observed that individuals receiving relatively large but yet sublethal doses of irradiation (as judged by proximity to the hypocenter in the absence of marked shielding) may or may not develop nausea and vomiting within a few hours of their exposure. There follows a relatively asymptomatic period of approximately two weeks, after which characteristic signs and symptoms appear. Some of these, such as fever, malaise, anorexia, nausea, and vomiting, have a low degree of specificity, being present in a wide variety of diseases. Other findings were much more specific. The chief among these are summarized in Table 4.1. These are figures for all survivors, irrespective of the amount of shielding which protected them. Let us restrict ourselves, for the moment, to a consideration of the findings in persons reporting themselves within 1,000 meters of the hypocenter. From the standpoint of distinctiveness, two symptoms —epilation and purpura—are outstanding. Oropharyngeal lesions also were frequently encountered. Diarrhea, while reported by large numbers of persons, could, at 20 days after the bombing, in any individual case be as well attributed to the poor hygienic conditions following the bombing as to the bomb itself, and does not seem to be a satisfactory symptom on which to base a study. Bloody diarrhea is a somewhat more reliable symptom, but still, from the findings of the Joint Commission, less useful than the other three symptoms already discussed.

Further evidence as to the relative validity of the findings listed in Table 4.1 as indicators of radiation exposure comes from an analysis of the manner in which the various findings tended to be associated. Epilation and petechiae were more highly associated than any other pair of symptoms in both cities. The next highest associations observed were, in both cities, epilation with oropharyngeal lesions, and petechiae with oropharyngeal lesions.

On the basis of these findings, it was felt that the appearance of epilation, petechiae, or oropharyngeal lesions, singly or in combination, within three months of the bombing, provided a relatively specific and objective yardstick of the absorption of a certain amount of radiation, such as could be employed in the present study (see also, inter alia, Warren and Bowers, 1950; Los Alamos Scientific Laboratory, 1950; vor der Bruegge, 1952; Hempelmann, Lisco, and Hoffman, 1952). Although exact data are lacking, it seems likely, on the basis of animal data, that there is considerable individual variation in susceptibility to these symptoms, a point which must of course be borne in mind when we come later to the problem of estimating radiation dosages. It will be noted that there are significant differences between Hiroshima and Nagasaki in the percentages of persons exhibiting certain symptoms subsequent to exposure within 1,500 meters of the hypocenter. These differences are customarily attributed primarily to local differences in shielding factors (see below), but may also be due in part to actual dose differences.

4.3.2The relation between distance from the hypocenter and radiation dosage.—Although the intensity of radiation obviously must decrease with distance from the hypocenter, the exact form of the distance-dosage relationship in Hiroshima and Nagasaki remains to this day uncertain. In part this is due to inadequate information concerning certain physical properties of these atomic explosions, in part to lack of detailed information concerning atmospheric moisture content at the time of the explosion. Furthermore, there is the possibility that the distance-dosage relationship is not the same in all radii from the hypocenter. Be this as it may, at the time this study was planned it was clear from such data as are given in Table 4.1, concerning the relationship between the occurrence of certain of the symptoms described above and distance from the hypocenter, that it would be important to record position at the time of the bombing as accurately as possible.

4.3.3The role of shielding in determining radiation dose.—The third type of information thought to be especially valuable in estimating radiation dosage concerned the amount of shielding protecting the individual from the effects of the explosion. Table 4.2 illustrates some of the findings of the Joint Commission in this respect. For the sake of brevity, only the Hiroshima data are reproduced; the findings in Nagasaki were essentially similar. An obvious question which had to be disposed of early in the design of the program concerned the elaborateness of the shielding data to be collected. Obtaining a complete shielding history can be quite time consuming. Theoretically, knowing the distance-dosage relationship and the precise shielding, one can arrive at a relatively accurate estimate of the amount of irradiation received by an individual. We have already mentioned the uncertainties surrounding the distance-dosage relationship. These uncertainties are particularly acute as regards the neutron component of the radiation spectrum, a component which in animal experiments has been shown to have a high relative biological effectiveness with respect to genetic effects. Furthermore, in the authors' opinion there is no realistic approach to the problem of radiation scatter and secondary radiation as a factor in the exact dosage received by any individual. Finally, there is room for reasonable doubt as to most individuals' abilities to reconstruct their exact positions at the time of so traumatic an event as this, and yet nothing less than such a reconstruction will suffice for a precise evaluation of shielding. With all these considerations in mind, it was concluded that while the experience of the Joint Commission left no doubt as to the desirability of collecting shielding data, this should be simple in nature.

4.4The type of radiation data collected in this study.—In addition to the biological considerations which we have just discussed, certain “practical” considerations entered into the determination of the type of data to be collected. The histories were to be obtained by clerks with the equivalent of a high school education. These clerks, although they could be carefully drilled in their duties, could not be expected to exercise a great deal of judgment if the approach to the problem involved the compilation of an elaborate history. It could be anticipated that the turn-over among clerks would be rather high. Finally, since the history taking was to extend over an indefinite number of years, much was to be said for concentrating on certain salient features of the individual's experience, which could be as readily recalled relatively late as relatively early.

With all these considerations in mind, the following information was obtained concerning each pregnant woman and her spouse included in this study:

1.

Presence in Hiroshima or Nagasaki at the time of the bombing.

2.

Location in city at time of bombing.

3.

Distance from the hypocenter [calculated from (2)].

4.

Indoors or outdoors.

5.

Type of building.

6.

Occurrence of subcutaneous bleeding (petechiae).

7.

Occurrence of gingivitis.

8.

Occurrence of bloody diarrhea.

9.

Occurrence of epilation (partial or complete).

10.

Occurrence of burns.

11.

Occurrence of external injuries.

In locating the position of an individual at the time of the bombing an effort was made to pinpoint position as accurately as possible. The distance from the hypocenter was then determined by actual measurement on a large-scale map of the city. The method of measurement has varied from time to time, involving variously the use of plastic and metal tapes, coordinates, and concentric circles. In the process of calculation, meters are rounded to decameters, and decameters further rounded to hectometers, so that the coded interval in hectometers denotes (x−55) to (x+44) meters.

Items 4 and 5 on the above list represent an attempt, however rough, to accumulate certain data relative to the evaluation of shielding. Item 4 was to be answered simply as indoors or outdoors. Item 5 pursued the point somewhat further, admitting of ten alternatives, namely, if indoors, then

1.

Inside concrete building,

2.

Inside brick building,

3.

Inside wooden Japanese or other type building,

4.

Inside cave or “bunker,” and if outdoors, then

5.

In open,

6.

Behind (within two meters) wall,

7.

In trench,

8.

Behind post or tree,

9.

In tram, train, or car, and

10.

Under eaves of house (i.e., shielded by house).

Items 6–9, bearing on radiation sickness, were to be answered simply as “yes” or “no,” after a brief explanation of the symptom by the clerk, if necessary. Limiting answers to “yes” or “no” obviously required the clerk to exercise her judgment in some cases. For the more vague term “oropharyngeal lesions” we have substituted gingivitis, since a necrotic-type of gingivitis with gingival bleeding was one of the more prominent and specific of the oropharyngeal lesions.

In the original design of the study, no symptom among those listed was to be recognized unless it developed prior to December 15, 1945. Some instances have come to light where the alleged onset of symptoms after that date have been recorded. This cannot be attributed to exposure to the bomb. There is a good reason to believe that the erroneous inclusion of such reports of “delayed effects” occurred very rarely.

Items 10 and 11 were included as one of several approaches to the problem of disentangling the possible radiation effects of the bombs from other effects and also the general disaster effect. Thus, it was anticipated that in the event of a positive correlation between some aspect of pregnancy termination and radiation history, an attempt could be made to determine if the correlation still held up after the elimination of those recording burns or external injuries of any type.

Those familiar with the problem of evaluating radiation exposure will be quick to recognize the very simple nature of the data collected in this study. The adequacy of these data as a basis for quantitative estimates of irradiation has been the subject of considerable discussion. Specifically, the possible desirability of a much more elaborate history was repeatedly discussed. It is our contention, admittedly subjective but documented in part by the internal consistency of the data presented in the following sections, the recall factor in these situations being what it is, and in view of the radiation variables, that this represents the most practical approach to this problem which could be adopted.

4.5The relation between distance, shielding, and symptoms in these data.—Tables 4.3 to 4.6 summarize the findings of the present study as regards the relationship between certain types of shielding, certain symptoms, and distance from the hypocenter. Based on items of the questionnaire (see preceding section), two degrees of shielding have been recognized, namely, that provided by being inside a Japanese-type home at the time of the explosion, and that provided by being inside any other type of structure—concrete or brick building, cave or bunker-type air raid shelter. Figures 4.1 and 4.2 present these same findings in graphic form. For ease of presentation, a “symptom ratio” has been used, consisting of the total number of symptoms (epilation, petechiae, or gingivitis) recorded by the survivors at a particular distance, divided by the number of survivors giving information. Data are not plotted for the 00–04 distances because of the few individuals involved. Several important conclusions emerge from a consideration of these data:

4.5.1 For all three groups, the symptom ratio falls off rapidly with distance. In the light of these findings, as well as the distance-dosage relationship estimated on physical grounds (see below), it seems unlikely that there was significant radiation beyond 3,000 meters from the hypocenter. The few individuals reporting epilation, petechiae, and/or gingivitis beyond that distance are, with possible rare exceptions, almost certainly in error in attributing their symptoms to irradiation. Taking this 1–2 per cent as a baseline for “false reporting,” it follows that the symptomatic group within 3,000 meters may also be somewhat “diluted” by false reporting.

4.5.2 There is evidence that shielding in “other style” buildings substantially reduced the proportion of persons with symptoms at distances within 1,500 meters.

4.5.3 In what appears at first sight to be a paradox, in the 500–900 meter ring a substantially higher proportion of persons in Japanese homes developed symptoms than did persons in the open. The reason for this seems clear. At this distance, persons reporting themselves in the open must actually have been shielded by one or more buildings from the bomb burst, otherwise they almost certainly would have sustained fatal burns. The biological evidence suggests that the total average shielding of such persons actually exceeded that of persons in Japanese homes.

4.5.4 In the 1,000–1,900 meter ring, there is a suggestion that in Hiroshima, although not so clearly in Nagasaki, presence in a Japanese-style home conferred some protection against the development of symptoms, the basis for comparison being persons reporting themselves “outdoors.” The average amount of shielding by which a person “in the open” was protected is very difficult to estimate. It is somewhat easier to quantitate the shielding of the persons in Japanese-style buildings. Woodbury (1953 and unpublished) reports that X-ray photographs taken at 62-kv indicate that the absorption of the bamboo-lattice and mud walls of the average Japanese house is equal to about 33 mm. of aluminum in the densest portion, and 15 mm. in the lightest. The absorptive value of the tile roofs which are standard construction in Japanese houses is undoubtedly more variable, depending on the distance from the hypocenter and the consequent variation in the angle of incidence of the radiation. However, for radiation incident at right angles, the average tile roof has an absorptive value in the neighborhood of 45 mm. of aluminum. Beyond 1,000 meters, persons in Japanese-type buildings in Hiroshima at the time of the explosion probably averaged several such walls or roofs between themselves and the exploding bomb. This is a significant degree of shielding. In very round terms, such shielding would screen out in the neighborhood of 50 per cent of the dosage at a level of 100r of the high-energy X-ray and gamma radiation released by the bomb.

It seems worth pointing out that the amount of shielding implied in the statement, “in a Japanese-style home,” varies directly with distance from hypocenter, since at greater distances, because of the angle of incidence of the radiation, more walls and roofs would usually intervene between the source of radiation than at lesser distances. Persons in Nagasaki in Japanese homes at the time of the explosion on the average were probably less shielded than in Hiroshima, because of the difference in terrain and the extension of homes up the sides of the valley.

4.6Factors contributing to the validity of the radiation histories.—The value of several different persons obtaining a history from a given individual is well known in clinical medicine. The changes in emphasis and the new material introduced, as the patient passes from one interviewer to the next, are sometimes remarkable. From the standpoint of verifying information, the Genetics Program has been fortunate for the following reasons:

4.6.1 Over the 6-year span covered by this program, many women registered two or more times (cf. Sec. 5.6). At each registration, a history was obtained without reference to any previous history.

4.6.2 In 1950 a Radiation Census was carried out in both Hiroshima and Nagasaki. This census obtained brief data on each survivor of the bombings then living in Hiroshima and Nagasaki, including position at the time of the bombing.

4.6.3 Some of the parents falling within the scope of the Genetics Program—especially the more heavily irradiated—also came under scrutiny in connection with other special studies of the ABCC, such as the Adult Medical Program, the Ophthalmology Program, etc. Each of these latter programs obtained a very detailed radiation history.

All of the information available on any given person was correlated through a Master File. Particular attention was directed towards a comparison of the radiation histories obtained from a given person on different occasions. In the event of a discrepancy, an effort was made to query the individual concerned, in an attempt to resolve it. On the basis of the information obtained on the query, a single uniform entry was made on all the forms involved. As might be expected, changes on the Genetics Short Form were not infrequent, although the data have not been kept in such a fashion as to permit a precise analysis. However, it is our impression that important changes were not common.

4.7Definition of radiation categories.—We are now in a position to define certain “radiation categories” which will be basic to the analysis which follows. This is done in Table 4.7. “Heavy” shielding denotes presence in concrete or brick building or air raid shelter at the time of the explosion. “Moderate” shielding includes being within a street car, train, or car, behind a wall or under the eaves of a house, i.e., on the side of the house away from the hypocenter. Finally, “light” shielding includes those giving their location as in a Japanese-style building or in a trench or behind a post or tree. Trench-type shielding is classified as “light” because of the difficulty in establishing how well an individual was protected by a trench.

Inspection of Table 4.7 in the light of the preceding discussion suggests that this is a somewhat “conservative” classification, in terms of correction for shielding. Thus, individuals reporting “heavy” shielding but asymptomatic have all been relegated to group 2, despite the fact that the symptom ratio in heavily shielded persons within 1,000 meters is approximately 0.7. Since relatively few persons were “heavily” shielded (see below), the net effect of this will be to increase slightly the average radiation dosage of group 2. On the other hand, this removes from categories 3 and 4, which received intermediate amounts of radiation, a group of individuals whose average radiation dose would be extremely difficult to evaluate.

The distribution of all the parents of registered terminations included in this study in terms of their radiation category is given in Table 4.8. Approximately 78.3 per cent of all the fathers who were registered, and 57.0 per cent of all the mothers, were not in the cities at all at the time of the explosion. Of those present in the cities, a substantial majority was beyond 3,000 meters from the hypocenter. Only 1.2 per cent of fathers, and 2.4 per cent of mothers, reported symptoms characteristic of radiation sickness.

In a preliminary report on this study (Neel et al., 1953), a system of radiation categories was employed which failed to take into account shielding. Thus, in this treatment group 4 was defined simply as: “In one of the two cities, and less than 1,845 meters from the hypocenter, but asymptomatic. Most of these individuals were shielded to a greater or lesser extent from the full effects of the bombs.” In this preliminary treatment, for some purposes the offspring of parents both of whom were group 4 or 5 were considered. In the case of the malformation analysis, for example, there were available 596 infants born to parents meeting these restrictions. But with the radiation categories which have just been defined, there are in the malformation analysis (Table 8.4) only 145 infants born to parents both in categories 4 or 5. The difference in numbers is due almost entirely to the attempt to correct for shielding.

4.8Considerations in the estimation of the average amount of radiation received by persons in each of the five radiation categories.—We come now to an attempt to estimate the average amount of radiation represented by each of the five radiation categories. It should at once be made clear that with the possible exceptions of categories 1 and 5, any estimate is at best an exceedingly rough approximation. For categories 2, 3, and 4, we will do well if we establish the range within which the true average is likely to be found. A particular difficulty in estimating radiation dosage, already alluded to, is the fact that a wide spectrum of radiant energy is involved, with the consequent complications in expressing tissue dosage in simple terms, a difficulty only partially circumvented by expressing dosage in terms of the “roentgen equivalent physical” (rep) unit. This latter may be defined as that amount of any ionizing radiation (beta rays, neutrons, etc.) which dissipates the same amount of energy per gram of water as one roentgen of X-rays or gamma rays. All estimates refer to surface dose. The actual amount of irradiation delivered to the gonad will be less, although because of the high energy content of atomic bomb irradiation, the gonad dose will actually be rather close to the skin dose.

Category 1 individuals, outside the cities at the time of the bombing, of course have received no radiation from the atomic bomb (ignoring, of course, the possibility of residual radiation). Category 5 individuals, symptomatic and within 3,000 meters, are considered to have received as a minimum the equivalent of 200 roentgens of gamma rays (Los Alamos Scientific Laboratory, 1950). The maximum dose received by a category 5 individual is in the neighborhood of the equivalent of 600 roentgens of gamma rays. The latter figure is based on the premise that the LD₅₀ in man for whole-body radiation is about 400 roentgens of gamma rays, but that a few individuals may survive doses amounting to 600r of gamma rays (Hempelmann, Lisco, and Hoffman, 1952). The average dose received by category 5 individuals can probably be placed at the equivalent of 300–400 roentgens of gamma rays. We will return shortly to the important question of the magnitude of the neutron component of this estimated dosage. Suffice it to say at this point that category 5 persons as a group received a relatively larger amount of neutron radiation than any other group, and because of the relatively greater biological effectiveness of neutrons as contrasted to gamma rays, generally speaking, the dose in rep units received by category 5 persons may be substantially below the dose as expressed in roentgens.

In arriving at our estimate of the amount of radiation received by category 2, 3, and 4 individuals, three considerations are outstanding:

4.8.1The estimated distance-dosage curve. —Figures 4.3 and 4.4 are the distance-dosage curves in terms of neutron and gamma radiation for the explosion of a “nominal” atomic bomb, roughly comparable to what was detonated over Hiroshima and Nagasaki, as published in “The Effects of Atomic Weapons.” In the absence of other unclassified data, this must of necessity serve as the only guide available. Note that the abscissa in both figures is scaled in terms of feet rather than the meters in which we have expressed distance. Figure 4.3, if taken at face value, indicates that beyond approximately 1,600 meters the amount of gamma radiation received was less than 100r. Our own observations, involving the reported appearance of symptoms in, e.g., significant numbers of persons in the 2,000–2,400 meter ring, lead us to feel this estimate is somewhat conservative. Also, unpublished calculations of Wilson (1951) suggest that, at least for Hiroshima, the neutron curve given in Figure 4.4 is on the conservative side. On the other hand, both from these two curves and on the basis of our own observations regarding the shape of the distance-dosage curve, it seems likely that persons at distances in excess of 3,000 meters received little if any radiation.

4.8.2The observations of the Joint Commission regarding leucopenia.—One of the well recognized effects of whole-body irradiation in amounts compatible with survival is a temporary leucopenia. Information concerning the proportion of individuals in the radiation categories defined above who developed leucopenia following their exposure would therefore be of value in any attempt to estimate radiation dosage. Certain of the published observations of the Joint Commission are quite useful in this respect. The more pertinent portion of their published data is summarized in Table 4.9 (see also LeRoy, 1950). Leucopenia is here defined as a total leucocyte count per mm.³ of less than 3,000 at some time during the second through the fifth weeks following the bombing. Some of the individuals on whom the percentages in Table 4.9 are based had repeated leucocyte determinations during a hospitalization, others had but a single count, at any time between the second and fifth weeks. The figures are thus at best estimates of the percentage which on careful studies would be found to develop leucopenia. The exposure categories utilized by the Joint Commission and reproduced in Table 4.9 are somewhat different from our own, the four most severe, the ones with which we are concerned, being defined as shown in Table 4.10.

It will be noted that the figures regarding leucopenia are based on fewer individuals than those with respect to epilation and petechiae. This is understandable in view of the difficulties in obtaining leucocyte determinations imposed by the conditions prevailing following the bombing. As the Commission report points out, these figures must be accepted with some reservations in view of the unfavorable circ*mstances under which they were obtained; on the other hand, the manner in which the findings regarding leucopenia parallel those on petechiae and epilation gives indirect confirmation of their validity. Since the Commission worked with a selected sample, composed largely of injured and, in the majority of cases, hospitalized survivors, caution must be exercised in extrapolating to our own, total sample. It will be noted, however, that with the exception of exposure category A, more of the persons examined developed leucopenia than either epilation or petechiae.

Through the courtesy of The Armed Forces Institute of Pathology, it has been possible to obtain certain very useful additional tabulations of the material collected by the Joint Commission. Among exposed Japanese in Hiroshima who received one or more leucocyte determinations sometime between the second and fourteenth weeks, who failed to develop epilation, petechiae, or gingivitis, and who are not known to have died within the fourteen weeks following the bombings, the distribution of lowest observed leucocyte counts by distance is as shown in Table 4.11. More intensive studies would have shown a higher proportion with leucopenia. The concentration of values in the “9,600 and above” class would seem to indicate a grouping of values which would vitiate attempts to calculate a mean. However, generally speaking, these observations suggest that a significant number of persons in our radiation categories 3 and 4 probably developed leucopenia as that term is here defined. An exact estimate of the proportion is difficult because of the difference between the way the Joint Commission's and our own data are broken down, but in round figures, based in part on the data in Table 4.11, it can be estimated that some 5–10 per cent of group 4 parents and 2–4 per cent of group 3 parents developed leucopenia.

The average amount of whole-body irradiation which, when delivered over a period of a day or two, will produce in man leucopenia of this degree is not known with certainty. For obvious reasons, there are not many pertinent observations. Thus, studies of patients with leukemia or multiple myeloma (e.g., Collins and Loeffler, 1956), while valuable for therapeutic reasons, obviously do not permit extrapolation to normal persons. Of the more recent investigations, the most valuable for present purposes—although still, because of the nature of the subjects, to be interpreted with caution— are those of Nickson (1951), who found that in three patients with carcinoma with metastases, 120r of whole-body irradiation with 400-kv X-rays produced no significant leucopenia. Two of these patients had a moderate leucocytosis at the time of treatment, amounting to 14,700/mm.³ and 13,900/mm.³, respectively. A fourth patient, also with carcinomatosis, who was treated with 107r of 200-kv whole-body X-rays on two successive days likewise failed to develop significant leucopenia. The impropriety of extrapolating from such patients to normal individuals is apparent. Furthermore, it does not appear that these patients were followed long enough for the detection of late leucopenic effects. The studies of Hempelmann, Lisco, and Hoffman (1952) on the few persons who have been involved in industrial radiation accidents are also especially pertinent. In one case, an individual was exposed to an amount of irradiation estimated as the equivalent of 186 roentgens of 80-kv X-rays and 10.7 roentgens of γ-rays, and in another case, to the estimated equivalent of 140r of 80-kv X-rays and 8.7r of γ-rays. In neither case did leucopenia as here defined develop.

In March of 1954 certain natives of the Marshall Islands and a lesser number of American military personnel were accidentally exposed to the fall-out of radioactive materials subsequent to the explosion of an experimental thermonuclear device. Detailed hematological studies were carried out on these individuals (Cronkite, Bond, and Dunham, in manuscript). Through the courtesy of Dr. E.P.Cronkite and Mr. Hyman Hechter, the original data on the Marshall Islanders regarding total leucocyte counts during a 10 to 70-day period following the exposure have been made available to us. There were two groups of Marshall Islanders involved, one group of 64 estimated to have received an average dose of 175r of whole-body, gamma radiation as measured in air, and a second group of 18 estimated to have received an average dose of 69r of gamma radiation as measured in air. The actual dose received by the Marshallese has been the subject of lively discussion. The observed maximum depression in the total leucocyte count occurred between days 39 and 51 following the exposure (counts were made every four days). On day 39, 3.1 per cent of the Group I Marshallese were found to have a leucopenia of 3,000 WBC/mm.³ or less; on day 43, 6.5 per cent; on day 47, 3.1 per cent; and on day 51, 0.0 per cent. None of the Group II natives showed a leucopenia of this degree during this period.

A comparison of these figures with the Japanese figures cited above is biased by the fact that the Marshall Island figures refer to the period of peak depression, whereas the Japanese data include a longer period of time. Nevertheless, the important fact emerges that in a group of healthy persons known with some accuracy to have received a dose of whole-body gamma irradiation of approximately 175r, the frequency of leucopenia was probably no greater than (if as great as) that occurring in the category 4 parents of this study. The obvious inference is that the average radiation dose experienced by category 4 parents is at least the equivalent of 200r of whole-body gamma radiation, with category 3 parents receiving a smaller average dose, but very likely one in the neighborhood of the equivalent of 100r of whole-body gamma radiation.

4.8.3The proportion of individuals protected by various types of shielding.—The final consideration to be introduced into this attempt to approximate the average radiation exposure in the various categories is the proportion of individuals experiencing various types of shielding. Data are not available for the whole sample. However, in a special study carried out in 1953 on all parents falling in the 1,800–2,500 meter ring, a group especially critical to this study, the breakdown shown in Table 4.12 was observed for 4,515 Hiroshima parents and 2,083 Nagasaki parents. The important point which emerges is the relatively small proportion of persons classified as receiving heavy or moderate shielding. It will be recalled that in the radiation classification adopted here, the report of “heavy” or “moderate” shielding resulted in an individual being placed in a lower radiation category. The contribution of such individuals—whose shielding is especially difficult to evaluate—to the composition of any radiation category is small.

4.9Estimates of the average amount of irradiation received by individuals in the various exposure categories.—We are now in a position to approximate the average amount of whole-body irradiation received by individuals in categories 2 through 5. Category 2 individuals, on the basis of the distance-dosage curve, probably received on the average no more than 5–10 roentgens equivalent physical (rep's). Category 3 survivors, on the same basis, as well as the known incidence of leucopenia and the shielding data, may be estimated on the average to have received 50–100 rep's. The estimation of the dosage in rep's for groups 4 and 5 is more difficult because, in contrast to groups 2 and 3, there is, on the basis of the distance-dosage curves and the occurrence of radiation cataracts in appreciable numbers of these survivors, a significant neutron component in the irradiation experienced by these persons. To make approximate allowance for this, we will express “the equivalent of 300–400 roentgens of gamma rays” suggested earlier as the average dose in category 5, as 200–300 rep's, and on the same grounds estimate the average dosage for category 4 individuals as 100–150 rep's. The arbitrary nature of these estimates will be apparent to all students of radiobiology. Any one might well be off by a factor of 2. We feel that any attempt to be more specific cannot be justified. In arriving at what some will term overly conservative estimates, we have been influenced by the fact that traumatic and thermal injuries may in a significant proportion of persons have combined with ordinarily sublethal radiation effects in causing death, thus lowering the mean radiation dose expected in the various categories on radiological grounds alone. On the other hand, any allowance for residual irradiation would revise the dose upward.

There are at present no published data which permit making a realistic estimate of the contribution of fast and slow neutrons to the total radiation spectrum. This is unfortunate in view of the relatively high biological efficiency of neutrons in inducing genetic change (summary in Symposium: Some Biological Effects of Radiation from Nuclear Detonation, 1954). However, it is important to recognize that the average dosages estimated above include, in the case of category 4 and 5 persons, a component of high effectiveness in producing genetic change.

Even should the upper rather than the lower estimates of the average amount of radiation received by survivors prove correct, the fact is clear that in the terms of the radiation geneticist, these are small doses indeed. This was recognized from the outset. But despite the improbability of being able to detect induced genetic effects in this material, there was no doubt in the minds of those who considered the problem initially, at all echelons, that “this unique possibility for demonstrating genetic effects caused by atomic radiation should not be lost” (cf. p. 2).

Note added in proof.—At the time this chapter was written, only the distance-dosage curves for a “nominal” atomic bomb had been published. However, since this manuscript went to press, distance-dosage curves, in terms of neutron and gamma radiation, have been declassified for both Hiroshima and Nagasaki. These are presented in Fig. 4.5. The gamma curves for the two cities are quite comparable, and correspond satisfactorily with those published earlier for the explosion of a “nominal” atomic bomb. However, it will be noted that the neutron curves for the two cities appear to differ significantly, with, at a given distance, rather more radiation of this type in Hiroshima than in Nagasaki. The curve for the latter city approximates that published earlier for the explosion of a “nominal” bomb. In view of the many uncertainties that enter into assigning mean dosages to our radiation exposure categories, it is not felt that these new data warrant a revision of the estimates given in this chapter. On the other hand, it is clear that there is a very strong neutron component in the radiation experienced by category 4 and 5 parents in Hiroshima, and in this connection, it will be recalled that approximately 68 per cent of all category 4 or 5 parents were from Hiroshima. In view of the high relative effectiveness of neutrons in inducing genetic change, the publication of these new curves has the effect of increasing the genetic significance to be attached to the observations herein to be reported.