Does radon cause diseases other than lung cancer? Findings on mortality within the German uranium miners cohort study, 1946-2018

Abstract

Objectives: This analysis aimed at investigating the relationship between death from diseases other than lung cancer and radon exposure in the German uranium miners cohort study with follow-up data from 1946 to 2018.

Methods: The cohort includes 58 975 men who were employed between 1946 and 1990 at the former 'Wismut' uranium mining company in Eastern Germany. Occupational exposure to radon progeny in working level month (WLM) was retrospectively assessed using a comprehensive job-exposure matrix. Based on internal Poisson regression, excess relative rates (ERRs) per 100 WLM were estimated for cumulative lagged exposure to radon for numerous outcomes. Sensitivity analyses examined possible confounding by occupational exposure to silica dust and gamma radiation.

Results: Small but statistically significantly increased excess rates were found for the group of all cancers other than lung cancer (n=6126; ERR/100 WLM=0.014 (95% CI 0.007; 0.022)) and for ischaemic heart diseases (n=6182; 0.010 (95% CI 0.003; 0.016)). The increase in risk was particularly observed at very high exposure levels. No clear association between radon exposure and other causes of death, comprising other subgroups of circulatory system diseases, non-malignant respiratory diseases excluding pneumoconiosis, neurodegenerative diseases and the many considered individual cancer sites was present. Notable were the estimates for myeloid leukaemia (n=114; 0.076 (95% CI -0.011; 0.164)) and pharynx cancer (n=112; 0.070 (95% CI -0.041; 0.182)).

Conclusions: The findings of the Wismut cohort indicate small increased risks for a few selected outcomes. Overall, the study does not provide convincing evidence for an increased risk for other diseases than lung cancer due to radon.

Keywords: Epidemiology; Radiation, Ionizing; Radon; Risk assessment.

Figure 1. Comparison of relative rate estimates from linear and categorical model for frequent outcomes (main groups with more than 1700 deaths and subgroups with more than 2500 deaths) in relation to cumulative radon exposure. ERR/100 WLM estimate from simple linear model (1) displayed by black line with CI in grey. Relative rate estimates from model (2) with exposure in categories (0, >0 to <50, 50 to <100, 100 to <500, 500 to <1000, 1000 to <1500, 1500+ WLM) displayed by black dots with CIs as vertical bars. Dashed orange horizontal lines corresponding to ‘no association’ for comparison. NMRD, non-malignant respiratory diseases; WLM, working level month.