Cancer Incidence in a Cohort of Swedish Chimney Sweeps, 1958-2006

Objectives. We examined cancer incidence in an expanded cohort of Swedish chimney sweeps. Methods. We added male chimney sweep trade union members (1981-2006) to an earlier cohort (employed 1918-1980) and linked them to nationwide registers of cancer, causes of deaths, and total population. The total cohort (n = 6320) was followed from 1958 through 2006. We estimated standardized incidence ratios (SIRs) using the male Swedish population as reference. We estimated exposure as years of employment and analyzed for exposure-response associations by Poisson regression. Results. A total of 813 primary cancers were observed versus 626 expected (SIR = 1.30; 95% confidence interval = 1.21, 1.39). As in a previous follow-up, SIRs were significantly increased for cancer of the esophagus, liver, lung, bladder, and all hematopoietic cancer. New findings included significantly elevated SIRs for cancer of the colon, pleura, adenocarcinoma of the lung, and at unspecified sites. Total cancer and bladder cancer demonstrated positive exposure-response associations. Conclusions. Exposure to soot and asbestos are likely causes of the observed cancer excesses, with contributions from adverse lifestyle factors. Preventive actions to control work exposures and promote healthier lifestyles are an important priority.