Objectives ‘Dusty occupations’ and exposure to low-dose radiation have been suggested as potential risk factors for stomach cancer. Data from the German uranium miner cohort study are used to further evaluate this topic.

Methods The cohort includes 58 677 miners with complete information on occupational exposure to dust, arsenic and radiation dose based on a detailed job-exposure matrix. A total of 592 stomach cancer deaths occurred in the follow-up period from 1946 to 2003. A Poisson regression model stratified by age and calendar year was used to calculate the excess relative risk (ERR) per unit of cumulative exposure to fine dust or from cumulative absorbed dose to stomach from α or low-LET (low linear energy transfer) radiation. For arsenic exposure, a binary quadratic model was applied.

Results After adjustment for each of the three other variables, a statistically non-significant linear relationship was observed for absorbed dose from low-LET radiation (ERR/Gy=0.30, 95% CI −1.26 to 1.87), α radiation (ERR/Gy=22.5, 95% CI −26.5 to 71.5) and fine dust (ERR/dust-year=0.0012, 95% CI −0.0020 to 0.0043). The relationship between stomach cancer and arsenic exposure was non-linear with a 2.1-fold higher RR (95% CI 0.9 to 3.3) in the exposure category above 500 compared with 0 dust-years.

Conclusion Positive statistically non-significant relationships between stomach cancer and arsenic dust, fine dust and absorbed dose from α and low-LET radiation were found. Overall, low statistical power due to low doses from radiation and dust are of concern.