2020 年 55 巻 1 号 p. 32-39
The health effects of low-dose or low-dose rate of radiation are still controversial. There is little evidence to show if radiation risk is greater than other risks, such as lifestyle or socio-economic factors, including smoking. This study aimed to directly compare radiation and smoking risk on cancer mortality by deriving both risk factors simultaneously from one cohort. The study population was Japanese nuclear workers who were engaged until the end of March, 1999. A lifestyle questionnaire was distributed in 1997 and 2003 and smoking information was obtained. Radiation dose was supplied by Radiation Dose Registration Center in Radiation Effects Association. Poisson regression was used to derive radiation excess relative risk (ERR) per 100 mSv and smoking (ERR) per 20 pack-years. Radiation doses were lagged by 10 years. There were 71,733 subjects. The total person-years was 591,000, and the number of deaths for all cancers excluding leukemia was 1,326. For all cancers excluding leukemia, the ERR of radiation per 100 mSv was 0.08 (90% CI: −0.08, 0.28), and the ERR of smoking per 20 pack-years was 0.57 (90% CI: 0.44, 0.73). In addition to all cancers excluding leukemia, stomach cancer, lung cancer, smoking-related cancers showed significantly smaller radiation ERRs than smoking ERRs. These results suggest that, even if a low-dose radiation risk existed, it was much smaller than smoking risk.
