Abstract
Background
The etiology of brain cancer is poorly understood. The only confirmed environmental risk factor is exposure to ionizing radiation. Because nuclear reactors emit ionizing radiation, we examined brain cancer incidence rates in the USA in relation to the presence of nuclear reactors per state.
Methods
Data on brain cancer incidence rates per state for Whites by sex for three age groups (all ages, 50 and older, and under 50) were obtained from cancer registries. The location, number, and type of nuclear reactor, i.e., power or research reactor, was obtained from public sources. We examined the association between these variables using multivariate linear regression and ANOVA.
Results
Brain cancer incidence rates were not associated with the number of nuclear power reactors. Conversely, incidence rates per state increased with the number of nuclear research reactors. This was significant for both sexes combined and for males in the ‘all ages’ category (β = 0.08, p = 0.0319 and β = 0.12, p = 0.0277, respectively), and for both sexes combined in the’50 and older’ category (β = 0.18, p = 0.0163). Brain cancer incidence rates for counties with research reactors were significantly higher than the corresponding rates for their states overall (p = 0.0140). These findings were not explicable by known confounders.
Conclusions
Brain cancer incidence rates are positively associated with the number of nuclear research reactors per state. These findings merit further exploration and suggest new opportunities for research in brain cancer epidemiology.
Authors' reply to Kundi's comments on de Vocht et al. “time trends (1998-2007) in brain cancer incidence rates in relation to mobile phone use in England”