Breast cancer incidence among female flight attendants: exposure-response analyses

Objective The aim of this study was to examine the association of breast cancer incidence with cosmic radiation dose and circadian rhythm disruption in a cohort of 6093 US female flight attendants. Methods The association of breast cancer risk with cumulative cosmic radiation dose, time spent working during the standard sleep interval, and time zones crossed (all lagged by ten years), adjusted for non-occupational breast cancer risk factors, was evaluated using Cox regression. Individual exposure estimates were derived from work history data and domicile-and era-specific exposure estimates. Breast cancers were identified from telephone interviews and state cancer registries, and covariate data were obtained from telephone interviews. Results Breast cancer incidence in the overall cohort was not associated with exposure. Positive associations in breast cancer incidence were observed with all three exposures only among the 884 women with parity of >= 3. Adjusted excess relative risks for women with parity of >= 3 were 1.6 [95% confidence interval (95% CI) 0.14-6.6], 0.99 (95% CI -0.04-4.3), and 1.5 (95% CI 0.14-6.2) per 10 mGy, per 2000 hours spent working in the standard sleep interval, and per 4600 time zones crossed (the approximate means of the fourth exposure quintiles among breast cancer cases), respectively. Conclusions Positive exposure-response relations, although observed only in a small subset of the cohort, were robust. Future studies of breast cancer incidence among other workers with circadian rhythm disruption should assess interaction with parity to see if our findings are confirmed.