Abstract

The cohort consisted of 15,908 men and women who worked for at least six months between 1948 and 1977 in 30 participating manufacturing plants in the reinforced plastics and composites industry. These workers were occupationally exposed to the working environment in the industry, which included exposure to styrene. Cause specific mortality analyses were performed based on the standardised mortality ratio (SMR) with the United States population as a comparison. No significant excess of cause specific mortality was found for the total cohort. Mortality from cancer was slightly less than expected (SMR = 88.1). For cancer of the respiratory system, a small non-significant excess was detected (SMR = 116.1). For lymphatic and haematopoietic cancer, a non-significant deficit was found (SMR = 73.3). The observed mortality from leukaemia was similar to that expected (five observed v 4.76 expected deaths). The plants with hot processes (injection moulding, centrifugal casting, compression moulding, continuous lamination, and pultrusion) experienced a significantly increased SMR (177.9) for respiratory cancer, which was more than twice that (78.3) for those with cold processes (resin mixing, lay up and spray up, bag moulding, and filament winding). As potential exposure to styrene from hot processes is considerably less than that from the cold processes, this finding could not be attributed to occupational exposures. A subsequent nested case-control study consisting of 40 cases of deaths from respiratory cancer was conducted. Further information on detailed work history, occupational exposures, and smoking history was collected. The case-control study did not show any significant association between respiratory cancer and direct exposure to styrene (contained in polyester resins), duration of exposure to styrene, the type of process (hot or cold), or whether a resin was used. A statistically significant association (relative risk = 7.33) was found between cigarette smoking and respiratory cancer among the study subjects.