Mortality and cancer incidence among secondary lead smelter workers.

OBJECTIVES To examine the mortality pattern and the cancer incidence in a cohort of long term smelter workers exposed to lead. METHODS The cohort consists of 664 male lead battery workers, employed for at least three months in 1942-87. From 1969 the values of all blood lead samples repeatedly obtained from these workers every two to three months, have been collected in a database. The expected mortality and morbidity 1969-89 was estimated from the county rates, specified for cause, sex, five-year age groups, and calendar year. Individual exposure matrices have been calculated and used for dose-response analyses. RESULTS The total cohort showed an increased overall mortality (standardised mortality ratio (SMR) 1.44; 95% confidence interval (95% CI) 1.16-1.79), an increased mortality from ischaemic heart diseases (SMR 1.72; 95% CI 1.20-2.42) and all malignant neoplasms (SMR 1.65; 95% CI 1.09-2.44). These risk estimates were unaffected or slightly decreased when applying a latency period of 15 years, and no dose-response pattern was shown. The non-significantly raised cancer incidence in the gastrointestinal tract (11 malignancies) in the total cohort, increased to a barely significant level in the quartile with the highest cumulative lead exposure (standardised incidence ratio (SIR) 2.34, 95% CI 1.07-4.45). No clear dose response pattern was evident when further subdividing the data into those first employed up to 1969 v those first employed after 1969 when the blood lead monitoring programme started. The risk estimate for malignancies in the gastrointestinal tract was not related to latency time. The cancer incidence was not increased at other sites. CONCLUSIONS A slightly increased incidence of gastrointestinal cancers was found in workers exposed to lead and employed before 1970. The lead cohort also showed an increased mortality from ischaemic heart diseases. These risk estimates did not show a dose-response pattern and were not associated with latency time. The results must also be interpreted with caution because of limited numbers, and lack of dietary and smoking data.