Radiation Effects on the Incidence of Myocardial Infarction in Atomic Bomb Survivors, 1958–2015

With the increasing use of radiation in medical and other settings, the potential effects of low-dose radiation on coronary heart disease including myocardial infarction (MI) are of great public health concern. The impact of low-dose radiation on myocardial infarction incidence remains controversial. The purpose of this study was to examine whether atomic bomb radiation exposure (dose < 4 Gy) is associated with the incidence of myocardial infarction. This prospective cohort study included 11,838 Japanese atomic bomb survivors with individually estimated radiation doses from the Adult Health Study cohort who had no history of myocardial infarction or radiotherapy at the first visit. Participants were followed until the earliest of first occurrence of an myocardial infarction, death, or the end of 2015 (57 years maximum). The incidence of myocardial infarctions (non-fatal and fatal), categorized by graded diagnostic accuracy, and the dose-response relationship of atomic bomb radiation were analyzed using Cox proportional hazards models. A total of 515 incident myocardial infarctions were documented, consisting of 188 definite, 30 probable, and 15 possible non-fatal myocardial infarctions, along with 282 fatal myocardial infarctions. For definite and probable non-fatal myocardial infarctions with high diagnostic accuracy, no significant association with radiation was found for both sexes combined [hazard ratio (HR) at 1 Gy = 1.17; 95% confidence interval (CI): 0.91-1.51], but sex-stratified analyses indicated a higher HR for females than for males [at 1 Gy, HRfemale = 1.42 (95% CI:1.02-1.98), HRmale = 1.02, 95% CI: 0.74-1.41], although the difference was not statistically significant (P = 0.13). No statistically significant modification of radiation effect was identified by city, age at exposure, attained age, time since exposure, smoking, or alcohol use. For all myocardial infarctions, including possible non-fatal myocardial infarction and fatal myocardial infarction with low diagnostic accuracy, the HRs at 1 Gy were 1.04 (95% CI: 0.86-1.25) for both sexes combined, 1.14 (95% CI: 0.88-1.47) for females, and 0.96 (95% CI: 0.75-1.22) for males. Findings from this long-term cohort study of atomic bomb survivors suggest an association between exposure to atomic bomb radiation and the incidence of myocardial infarction in females, but not in males. Further studies are necessary to clarify the reasons underlying the sex difference in dose-response relationship for myocardial infarction.