Context Cohorts of nuclear workers enable the assessment of cancer risk related to occupational radiation exposure. However workers are also exposed to environmental and medical radiation. Those non-occupational doses can be of the same order of magnitude than the occupational ones. This study aims to consider the potential for change when adding non-occupational doses to occupational doses in dose-response analyses.

Methods Analyses were carried out on the French nuclear workers cohort. The cohort includes nuclear workers who were hired between 1950 and 1994 by CEA, AREVA or EDF, worked at those companies for at least one year and who were badge-monitored for occupational external radiation exposure which was available through dosimetric recordings. The mortality follow-up extends until 2004. Vital statuses and causes of death were obtained from national registries. Environmental and medical exposure scenarios based on available knowledge were used to attribute non-occupational doses. The association between occupational and non-occupational radiation exposure and cancer mortality was quantified using Poisson regression methods.

Results The cohort was constituted of 59,004 workers who were followed-up for 25 years on average. The mean cumulative Hp(10) dose of occupational radiation was 18.4 mSv. At the end of follow-up, 6,310 deaths had occurred, including 2,552 cancer deaths. The putative cumulative environmental dose changed depending on the scenario but was of the order of 130 mSv. The impact of accounting for non-occupational exposure will be discussed for the different scenarios considered.

Conclusion The French cohort of nuclear workers allows for comprehensive analysis of the cancer risk of chronic exposure to low doses external radiation. Past analyses of the cohort found a non-significant association between occupational exposure to radiation and cancer mortality. The aim of this study is to analyse the impact of medical and environmental exposures on the estimated association between occupational radiation exposure and cancer risk.