Research plan for RERF study of life-span of A-bomb survivors, Hiroshima and Nagasaki
Background and purposeWith the emergence of the hypothesis of accelerated aging as a radiation effect, and some suggestion in preliminary studies by Woodbury et al that death rates among A-bomb survivors may exceed those among the nonexposed, it has become plain that there may be A-bomb effects that would be brought to light only by a large, carefully designed mortality study. Although systematic collection of death certificates for residents of Hiroshima and Nagasaki began in 1951, extending retroactively to deaths occurring in 1948, it was not until the report of the Francis Committee in 1955 that the potential contribution of a mortality investigation came to be generally recognized in Japan and in the USA and that a technically adequate proposal was outlined. Only a large mortality study holds promise for the detection of small differences in longevity or, conversely, for the demonstration that any effect upon longevity must be negligibly small. A program of physical and laboratory examinations is necessarily conducted on samples of a size that renders comparisons based on them relatively insensitive to small mortality effects, and is inevitably haunted by the specter of incomplete, and therefore possibly biased, follow-up.
According to tabulations of the supplementary schedules prepared in connection with the 1950 National Census here were then about 284,000 survivors of the bombs, 159,000 from the first bomb and 125,000 from the second. In Hiroshima itself there were 98,000 and in Nagasaki 97,000. However, within 2000 m of the hypocenters there were only an estimated 39,000 in both cities, and within 1000 m only 2000.
These estimates suggest the need for sampling the dosage or distance categories differentially, taking most or all individuals surviving exposure close to the hypocenters and some suitable fraction of those exposed at greater distances from the hypocenters. The solution proposed by the Francis Committee seems straightforward: Distance is subdivided into several zones from each of which a fixed number of patinets is drawn to equal the count of those in the zone closest to the hypocenter.
In March 1959 the investigation of nonexposed in the two cities was far advanced, but still incomplete. Eligible cases had been amassed in the number of 13,685 in Hiroshima and 8,653 in Nagasaki, and this work was regarded as about 60% complete.
Collection of data
The largest task in the entire project is the creation of the rosters, in connection with which the bulk of the underlying information is obtained on the independent variables. The remaining observations that are needed are those concerned with survival status and the estiamtion of radiation dose.
Information on fact of death
Place and date of death are transcribed from the Report of Death to the Koseki Register. For those who have their honseki addresses in Hiroshima City or Nagasaki City, cards are prepared beforehand and arranged in the same order as the Koseki Register.
In case it is not possible to locate the honseki address of sample members by checking the Koseki Register, field investigators visit the homes of sample members to check the correct honseki address and ascertain whether the individuals are living or dead.
Estimation of radiation dose
The radiation dose is the fundamental independent variable of which distance from hypocenter and radiation symptoms are but correlates. Estimation of dose requires reliable air dose curves, accurate shielding histories, and attenuation factors for various shielding materials. For both cities air dose curves and attenuation factors for Japanese building materials have been provided to ABCC by the US Atomic Energy Commission.
AnalysisEstimated radiation dose is, of course, the independent variable throughout the analysis, and yet it seems possible that acute symptoms may be of value as an ancillary independent variable. Since estiamted dose and symptoms are independently obtained they may have some reinforcing value in the analysis.
Data from the two cities will be analyzed in parallel and conclusions reached on the basis of their joint consideration. Any relationship that seems established by the data of one city, but contradictory to those of the other, will be considered merely suggestive.
Comparisons will not be made in one-sided fashion, as would be appropriate to test the notion that any real radiation would necessarily increase mortality, but will be two-sided in recognition of the possibility that those who survived the acute radiation injury were by that token selected for longer-than-average survival.