Technical Report No. 1-91
Mortality among the offspring (F1) of atomic bomb survivors, 1946-85
Yoshimoto Y, Schull WJ, Kato H, Neel JVEditor’s note: Publications based on this report were published in J Radiat Res (Tokyo) 32S:294-300, 1991, and J Radiat Res (Tokyo) 32:327-51, 1991
Summary
We compare deaths occurring in the years 1946-85 in a cohort of 31,159 children born to parents one or both of whom were exposed to the atomic bombing of Hiroshima or Nagasaki, and received a combined (i.e., joint) gonadal dose of 0.01 Sv or more, with deaths in a comparable control group, totaling 41,069 children. The average combined gonadal dose equivalent for the exposed parents was 0.435 Sv. Gonadal doses were calculated using the recently established DS86 system, supplemented by an ad hoc system for those children for whom a DS86 dose could not be computed for one or both parents. At the end of 1985, those members of the study groups born in 1946 had reached 39 years of age, whereas those born in the years 1966 through 1984 had not yet reached their 20th birthday. The mean age of living members of the cohorts was 28.8 years.
When a linear relative risk model is fitted to the data, no statistically significant increase in the risk of mortality attributable to diseases other than neoplasms is noted following parental exposure, the excess relative risk being 0.030 (plus or minus 0.046) per sievert based on the subset of individuals with DS86 doses, assuming the RBE of neutrons to be 20. For fatal cancer, in confirmation of an earlier report on cancer incidence below the age of 20 in this same group, again no statistically significant effect was observed as parental radiation dose increased. Finally, although the present method of analysis using Poisson regression and person-years at risk of death seems more appropriate now, particularly as the cohort ages—since earlier analyses of mortality in the F1 cohort have been based on a simple linear regression of the frequency of death on parental dose—this model was also fitted to the data used in the relative risk estimate to provide some continuity with the past. The results give an intercept of 0.0420 (plus or minus 0.0015) and a linear regression coefficient of 0.00169 (plus or minus 0.00157) per sievert. This leads to the calculation of a (statistically nonsignificant) excess relative risk of 0.040, in good agreement with the excess obtained by fitting the relative risk model. An analysis based on the full sample, using not only the DS86 dose group but also the ad hoc dose group, yields essentially the same result as the analysis restricted to the DS86 dose group.
We compare deaths occurring in the years 1946-85 in a cohort of 31,159 children born to parents one or both of whom were exposed to the atomic bombing of Hiroshima or Nagasaki, and received a combined (i.e., joint) gonadal dose of 0.01 Sv or more, with deaths in a comparable control group, totaling 41,069 children. The average combined gonadal dose equivalent for the exposed parents was 0.435 Sv. Gonadal doses were calculated using the recently established DS86 system, supplemented by an ad hoc system for those children for whom a DS86 dose could not be computed for one or both parents. At the end of 1985, those members of the study groups born in 1946 had reached 39 years of age, whereas those born in the years 1966 through 1984 had not yet reached their 20th birthday. The mean age of living members of the cohorts was 28.8 years.
When a linear relative risk model is fitted to the data, no statistically significant increase in the risk of mortality attributable to diseases other than neoplasms is noted following parental exposure, the excess relative risk being 0.030 (plus or minus 0.046) per sievert based on the subset of individuals with DS86 doses, assuming the RBE of neutrons to be 20. For fatal cancer, in confirmation of an earlier report on cancer incidence below the age of 20 in this same group, again no statistically significant effect was observed as parental radiation dose increased. Finally, although the present method of analysis using Poisson regression and person-years at risk of death seems more appropriate now, particularly as the cohort ages—since earlier analyses of mortality in the F1 cohort have been based on a simple linear regression of the frequency of death on parental dose—this model was also fitted to the data used in the relative risk estimate to provide some continuity with the past. The results give an intercept of 0.0420 (plus or minus 0.0015) and a linear regression coefficient of 0.00169 (plus or minus 0.00157) per sievert. This leads to the calculation of a (statistically nonsignificant) excess relative risk of 0.040, in good agreement with the excess obtained by fitting the relative risk model. An analysis based on the full sample, using not only the DS86 dose group but also the ad hoc dose group, yields essentially the same result as the analysis restricted to the DS86 dose group.