Our results add to the evidence that ethylene oxide is a human carcinogen.

We explored the risk of lymphoma and its most prevalent subtypes associated with occupational contact with livestock, and whether risk was modified by age at first contact, in 2,348 incident lymphoma cases and 2,462 controls who participated in the EPILYMPH case-control study. A detailed occupational history was collected in cases and controls, including working in a livestock farm, species of livestock, its approximate number and circumstances of contact. For each disease outcome, and each type of livestock, odds ratios (OR) and their 95% confidence intervals (95% CI) were calculated using unconditional logistic regression, adjusting for age, gender, education and center. Lymphoma risk (all subtypes combined) was not increased amongst those exposed to contact with any livestock (OR 5 1.0, 95% CI 0.8-1.2). Overall, we did not observe an association between occupational contact with livestock and risk of lymphoma (all types) and B-cell lymphoma. The risk of diffuse large B cell lyphoma (DLBCL) was significantly lower amongst subjects who started occupational contact with any species of livestock before or at age 12 (OR 5 0.5, 95% CI 0.2-0.9), but not at older ages. A significant heterogeneity in risk of B cell lymphoma by age at first contact was detected for contact with cattle, poultry and swine. Early occupational contact with livestock might be associated with a decrease in risk of B cell lymphoma.

Our results support the hypothesis that early-life exposure to pets, birds and particularly with chickens might be associated with a reduced risk of lymphoma.

cancer. We designed the study to overcome certain limitations of earlier epidemiology studies. Method The epidemiological and exposure assessment components of the study are coordinated by the University of Pittsburgh and the University of Illinois at Chicago, respectively. Our cohort comprises 12 manufacturing sites in the US and nine sites in Europe, and represents three companies, five countries (US, Austria, Germany, Sweden and UK) and multiple manufacturing processes and exposures. Statistical analyses will adjust external and internal mortality rate comparisons for potential co-exposures, including smoking histories obtained via a nested case-control study. The study will include separate and pooled analyses. Results Our data collection effort identified two additional US study sites and additional subjects in the German and Swedish sites. Accordingly, our originally projected cohort size of 21 000 subjects has increased to 35 508 (US-7005; Europe-33 508). Vital status tracing, cause of death determination and identification of subjects for the nested case-control study are ongoing. Conclusions Our study will enable country-specific and pooled analyses of mortality rates and exposure-response relationships among workers from 21 study sites and the opportunity to compare and contrast findings across countries, sites, companies and/ or manufacturing processes and exposures involved in this global industry. We will detail progress to date on the US and combined epidemiological component of the study.