ORIGINAL ARTICLE Occupational Exposures and Breast Cancer Among Women Textile Workers in Shanghai Roberta M. Ray,* Dao Li Gao,† Wenjin Li,* Karen J. Wernli,*‡ George Astrakianakis,§ Noah S. Seixas,§ Janice E. Camp,§ E. Dawn Fitzgibbons,* Ziding Feng,* David B. Thomas,*‡ and Harvey Checkoway*†§ Background: Breast cancer incidence rates have been increasing in China over the past 2 decades. Most studies have focused on reproductive, dietary, and genetic risk factors. Little is known about the contribution of occupational exposures. Methods: We conducted a case-cohort study within a cohort of female textile workers who had participated in a randomized trial of breast self-examination in Shanghai, China. We compared 1709 incident breast cancer cases with an age-stratified reference subco- hort (n  3155 noncases). Cox proportional hazards modeling, adapted for the case-cohort design, was used to estimate hazard ratios for breast cancer in relation to duration of employment in various job processes and duration of exposure to several agents. We also evaluated the associations of cotton dust and endotoxin with breast cancer. Results: Cumulative exposures to cotton dust and endotoxin dem- onstrated strong inverse gradients with breast cancer risk when exposures were lagged by 20 years (trend P-values 0.001). We did not observe consistent associations with exposures to electromag- netic fields, solvents, or other chemicals. Conclusion: Endotoxin or other components of cotton dust expo- sures may have reduced risks for breast cancer in this cohort, perhaps acting at early stages of carcinogenesis. Replication of these findings in other occupational settings with similar exposures will be needed to confirm or refute any hypothesis regarding protection against breast cancer. (Epidemiology 2007;18: 383–392) B reast cancer rates in China are considerably lower than in the United States and other developed countries, although incidence rates have increased markedly in recent years. 1 Menstrual and reproductive factors are estimated to explain 40 to 50% of breast cancer risk in Shanghai. 2 Changing dietary and other lifestyle practices may be responsible for some of the increased incidence. 3 Occupational and other environmental risk exposures may also have contributed to the rise in breast cancer occurrence in China, but have been less thoroughly investigated than other factors. Although women comprise a substantial and increasing proportion of the workforce worldwide, there are very few specific established occupational risk factors for female breast cancer. A noteworthy exception is ionizing radiation, for which there is strong evidence of associations in certain occupational groups. 4 There are varying levels of support for causal associations with occupations and specific workplace exposures. Among the more consistent findings are elevated risks among teachers, librarians, and administrative workers, which may reflect the influence of factors associated with high socioeconomic status. 5 There has been considerable interest in possible associations of breast cancer with occu- pational and nonoccupational (residential) exposures to elec- tromagnetic field (EMF) exposures, although the evidence is inconsistent. 6 Nonetheless, some recent studies suggest mod- estly elevated risks. 7–9 Suppression of melatonin by EMF is the presumed mechanism of mammary carcinogenesis for this exposure. Somewhat stronger associations for breast cancer risk have emerged from investigations of night shift work, ostensibly acting by a similar mechanism of melatonin sup- pression. 10,11 Results from studies of breast cancer and per- sistent organochlorine environmental chemicals, such as PCBs and DDT, have been mixed. 12,13 Some studies indicate elevated risks related to occupational exposures to solvents, and polycyclic aromatic hydrocarbons, 14,15 but these associ- ations have not been observed consistently. There is very little evidence that machining fluids are related to breast cancer risk. 16 Relatively sedentary jobs have been linked to moderate risk excesses. 17 The textile industry is one of the world’s largest em- ployers of women, and this industry entails exposures to numerous potential carcinogens, including solvents, dyes, EMF, formaldehyde, machining fluids, cotton, wool, syn- thetic fibers, and silk dusts, and silica. 18,19 Neither consis- tently elevated overall breast cancer risks nor associations Submitted 13 July 2006; accepted 19 December 2006. From the *Program in Epidemiology, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA; †Department of Epidemiology, Zhong Shan Hospital Cancer Center, Shanghai, China; ‡Department of Epidemiology, University of Washington, Seattle, WA; and §Department of Environmental and Occupational Health Sciences University of Washington, Seattle, WA. Supported by grant R01CA80180 from the US National Cancer Institute, National Institutes of Health, and US National Institute of Environmental Health Sciences Training Grant ES 07262. Correspondence: Roberta M. Ray, Fred Hutchinson Cancer Research Center, M4-A402, 1100 Fairview Avenue North, PO Box 19024, Seattle, WA 98109-1024. E-mail: rray@fhcrc.org. Copyright © 2007 by Lippincott Williams & Wilkins ISSN: 1044-3983/07/1803-0383 DOI: 10.1097/01.ede.0000259984.40934.ae Epidemiology • Volume 18, Number 3, May 2007 383