Confounding Effects of Lifestyle Factors in Cancer Risk Estimation for Occupational Radiation Exposure

Abstract

Background

The confounding effect of lifestyle factors is an important concern in occupational studies, particularly when the risk magnitude is relatively small. This study aimed to evaluate the potential confounding effects of lifestyle factors on the association between radiation exposure and cancer incidence.

Methods

Data from all Republic of Korean diagnostic medical radiation workers enrolled in the national dose registry were merged with cancer incidence records up to 2018. Excess relative risks (ERRs) for cancer were calculated using Poisson regression models to quantify the radiation dose-response relationship. Major lifestyle factors were imputed using multiple imputations by chained equations based on survey data. The confounding effects were assessed by comparing ERRs before and after adjustment for lifestyle factors.

Results

The baseline ERR for cancer incidence per Sievert was 0.44 (95% CI: -0.94, 1.83) after adjusting for attained age, sex, birth year, and employment duration. Further adjustment for lifestyle factors (smoking status, alcohol consumption, body mass index, physical exercise, sleep duration, and night shift work) did not substantially modify this risk coefficient, with change-in-estimate values ranging from 0% to 13.6%. Sensitivity analyses conducted with the survey-based cohort and sex-stratified analyses yielded consistent results.

Conclusion

Our study found little evidence of significant confounding effects from unmeasured lifestyle factors on cancer risk when basic registry data variables were adjusted among medical radiation workers. Further studies are warranted to investigate the impact of unmeasured and unknown confounders to improve the accuracy of radiation risk estimates.