A cohort mortality study among workers at a 1,3 butadiene facility.

This is a cohort mortality study of 2800 male workers employed at least 6 months between 1943 and 1996 at a 1,3-butadiene monomer production facility. Earlier analyses of mortality for this cohort found statistically significant deficits for the 'all causes of death' category, and a lower than expected mortality for most leading causes of death. Past analyses also showed a significant elevation for deaths from cancers of the lymphohematopoietic system that was mainly due to an increase in deaths from lymphosarcoma. The purpose of this update was to examine the patterns of mortality through the end of 1999, for five additional years of follow-up. Persons who had become eligible between the last cohort update and April 1996 were added. Cohort membership was closed after April 1996 due to the sale of the facility. A total of 1422 deaths through December 1999 were identified, giving over 200 more deaths than in the last report on this aging cohort. The standardized mortality ratio (SMR) for all causes of death was 89 [95% confidence interval (95% CI) =84-94], which is statistically significantly low, and that for all malignant neoplasms was 90 (95% CI=81-101). The SMR for all lymphohematopoietic cancers (LHC) was 141 (95% CI=105-186) and is statistically significant. The SMR for leukemia was 129 (95% CI=77-204) and for non-Hodgkin's lymphoma (NHL), 148 (95% CI=89-231). The LHC elevations again were found only in workers first employed before 1950 and, in the group with the highest potential for exposure to butadiene, the elevations were highest in the short-term workers. Survival analyses were performed for all LHC [International Classification of Diseases (ICD) codes 200-209], NHL (ICD codes 200, 202), and leukemia (ICD codes 204-207) using an estimate of cumulative butadiene exposure as a time-dependent explanatory variable defined as a combination of job exposure class, calendar time, and length of time in job. The relative risks for the above causes of death were essentially 1.0, suggesting that there was no increase in risk with increasing butadiene exposure.