Exposure to low-dose ionizing radiation and dementia mortality in Canadian nuclear power plant workers

Evidence suggests that ionizing radiation increases the risk of dementia, but data are limited for those chronically exposed to low doses (<100 millisieverts (mSv)). We addressed this gap by investigating associations between cumulative low-dose radiation exposure and dementia mortality among nuclear power plant (NPP) workers.

Our study cohort comprised Canadian NPP workers monitored for radiation exposure between 1950 and 2018, with mortality follow-up extending through 2020. Dementia deaths were identified through record linkage to national mortality data and based on underlying or contributing causes of death. Standardized mortality ratios (SMRs) were derived to compare mortality in the workers to the Canadian population. We used Poisson regression and fit linear excess relative risk models to describe the shape of the exposure-response relationship between cumulative radiation (lagged 10 years) and dementia mortality.

Among 97,250 NPP workers, there were 1,458,299 person-years of follow-up. The mean whole-body lifetime accumulated exposure at the end of follow-up was 9.8 mSv. In total, there were 235 deaths with dementia listed as the underlying cause, and 450 when the definition was expanded to include contributing causes. Relative to the Canadian population, NPP workers had a reduced mortality rate of dementia (SMR = 0.69; 95% CI: 0.63, 0.76). We found some evidence that exposure to ionizing radiation was associated with an increased risk of dementia mortality, but no statistically significant linear dose-response relationship (ERR/100 mSv = 0.010; 95% CI: -0.227, 0.247). Spline analyses suggested a non-linear dose-response relationship.

We found modest evidence that low-dose radiation exposure is associated with increased risks of dementia mortality. These findings should be interpreted cautiously, given that they differ from those found when we modelled dementia as an incident outcome in a cohort that included many of these same workers. Taken together, our findings highlight limitations of using death data to identify etiological risk factors for dementia.