Exposure–response assessment of cancer mortality in styrene-exposed boatbuilders

Abstract

ObjectivesTo improve exposure estimates and reexamine exposure–response relationships between cumulative styrene exposure and cancer mortality in a previously studied cohort of US boatbuilders exposed between 1959 and 1978 and followed through 2016.MethodsCumulative styrene exposure was estimated from work assignments and air-sampling data. Exposure–response relationships between styrene and select cancers were examined in Cox proportional hazards models matched on attained age, sex, race, birth cohort and employment duration. Models adjusted for socioeconomic status (SES). Exposures were lagged 10 years or by a period maximising the likelihood. HRs included 95% profile-likelihood CIs. Actuarial methods were used to estimate the styrene exposure corresponding to 10-4 extra lifetime risk.ResultsThe cohort (n= 5163) contributed 201 951 person-years. Exposures were right-skewed, with mean and median of 31 and 5.7 ppm-years, respectively. Positive, monotonic exposure–response associations were evident for leukaemia (HR at 50 ppm-years styrene = 1.46; 95% CI 1.04 to 1.97) and bladder cancer (HR at 50 ppm-years styrene =1.64; 95% CI 1.14 to 2.33). There was no evidence of confounding by SES. A working lifetime exposure to 0.05 ppm styrene corresponded to one extra leukaemia death per 10 000 workers.ConclusionsThe study contributes evidence of exposure–response associations between cumulative styrene exposure and cancer. Simple risk projections at current exposure levels indicate a need for formal risk assessment. Future recommendations on worker protection would benefit from additional research clarifying cancer risks from styrene exposure.