A cohort study among workers at a metal components manufacturing facility.

To follow-up on a proportional mortality study that showed significantly elevated proportional mortality ratios for lung cancer and a subgroup of nonmalignant respiratory diseases, we conducted a cohort mortality study (1950-1987) among 4627 employees of a metal components manufacturing facility. The findings of this study showed lower than expected mortality from all causes of death and all cancers. However, lung cancer mortality was significantly elevated (standardized mortality ratio = 131, 95% confidence interval (102-165), owing to elevated mortality among hourly workers (standardized mortality ratio = 153, 95% confidence interval 118-195). Hourly workers also showed a significantly elevated rate for the residual category "other nonmalignant respiratory disease" (standardized mortality ratio = 170, 95% confidence interval 110-251) and a significant deficit of leukemia (standardized mortality ratio = 16, 95% confidence interval 0-87). Analyses by duration of employment did not show significant trends for any cause of death. Smoking information was not available, but several indirect methods were used to estimate the potential confounding effect of smoking.