Objectives The US Occupational Safety and Health Administration (OSHA) recently proposed a permissible exposure limit of 0.2 µg/m³ for beryllium, based partly on extrapolated estimates of lung cancer risk from a pooled occupational cohort. The purpose of the present analysis was to evaluate whether cohort members exposed at lower levels to mainly insoluble forms of beryllium exhibit increased risk of lung cancer.

Methods We conducted Cox proportional hazards regression analyses among 75 lung cancer cases in age-based risk sets within two lower exposure plants in the pooled cohort followed from 1940 to 2005. We used categorical and power models to evaluate exposure–response patterns for mean and cumulative beryllium exposures in the two-plant cohort, comparing findings with the full pooled cohort. We also evaluated the distribution of exposure-years in each cohort by solubility class (soluble, insoluble and mixed).

Results 98% of workers in the two-plant cohort were hired between 1955 and 1969. The mean beryllium exposure averaged 1.3 µg/m³ and the predominant form was insoluble. Adjusting for confounders, we observed a monotonic increase in lung cancer mortality across exposure categories in the two-plant cohort. The exposure–response coefficients (per unit ln exposure) were 0.270 (p=0.061) for mean exposure and 0.170 (p=0.033) for cumulative exposure, compared with 0.155 and 0.094 (respectively) in the full cohort.

Conclusion The low-exposure levels at these two plants and the predominance of insoluble beryllium suggest that the overall pooled cohort findings on which OSHA’s lung cancer risk assessment is based are relevant for current workers exposed to any form of beryllium.