IJE Advance Access originally published online on September 15, 2006
International Journal of Epidemiology 2006 35(5):1359; doi:10.1093/ije/dyl203
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Letter to the Editor |
RE: ‘Occupational exposures and risks of liver cancer among Shanghai female textile workers—a case–cohort study’
Department of Community & Family Medicine, The Chinese University of Hong Kong, 4/F School of Public Health, Prince of Wales Hospital, Shatin, N.T., Hong Kong SAR, China
* Corresponding author. Department of Community & Family Medicine, The Chinese University of Hong Kong, 4/F School of Public Health, Prince of Wales Hospital, Shatin, N.T., Hong Kong SAR, China E-mail: iyu{at}cuhk.edu.hk
In a recent issue of the journal, Chang et al.1 reported a protective effect of cumulative exposure to endotoxin or cotton dust on the incidence of liver cancer among female textile workers after allowing for 20 years latency. On the other hand, an unanticipated elevated risk of liver cancer (Hazard ratio = 5.30 with 95% confidence interval: 1.29–21.81, based on three incident cases) was observed among female workers exposed to inks for less than 10 years, but the underlying reasons were not explored by Chang et al.
The authors failed to explore the potential confounding effects from hepatitis B virus (HBV) infection and dietary risk factors. This is an intrinsic methodological defect in this type of case–cohort analysis—inadequate exploration of major potential confounding factors for individual outcomes. Controlling for major potential confounding factors is more likely to be possible in a properly conducted nested case–control study. It was possible that workers known to be carriers of hepatitis B or C would be advised against taking up jobs that entailed exposure to endotoxins, and this could adequately explain the apparent protective effects of cotton dust and endotoxin exposures. In any case, exposing workers to endotoxins for their potential protective effect against cancer should not be encouraged, as endotoxins do have their harmful health effects on workers, including byssinosis and other diseases.
The non-significant inverse relationship between oral contraceptive and liver cancer could also be explained by a form of ‘survivor selection’ bias rather than simply due to a high prevalence of HBV infection among the population that contributed to a considerable portion of liver cancer incidence, as interpreted by Chang et al.1 Female workers with HBV infection might be more likely to use intrauterine contraceptive device to avoid liver function being further damaged by oral contraceptives.
The authors did not offer any explanations for the apparent excess risk of liver cancer among workers with <10 years of inks exposure or the apparent protective effect of exposures to electromagnetic field or non-ionizing radiation, but these could also have resulted from confounding by HBV infection and dietary aflatoxin contaminations. The findings of the study by Chang et al. would have been more informative if the potential confounding effects from both HBV infection and dietary aflatoxin were taken into consideration.
Another caveat of the study was that the authors lumped primary malignant neoplasm of liver (code: 155, ICD-9 revision) and malignant neoplasm of liver, not specified as primary or secondary (code: 155.2), together as the outcome of liver cancer. To strengthen their conclusions, we strongly recommend the authors to perform a sensitivity analysis by limiting the statistical analyses to the primary liver cancers (code: 155, ICD-9 revision) and check whether there are significant divergences from the original findings.
Conflict of interest: None declared.
Reference
1 Chang CK, Astrakianakis G, Thomas DB et al. Occupational exposures and risks of liver cancer among Shanghai female textile workers—a case–cohort study. Int J Epidemiol 2006;35:361–69.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||