Document resource
Objective To conduct an updated systematic review and meta-analyses of the exposure-response relationship between cumulative occupational asbestos exposure and the risk of lung cancer.Material and Methods The search strategy was executed on five databases and by review of reference lists. A total of 32 cohort and case-control studies with risk estimates for cumulative, categorical asbestos exposure and lung cancer risk were included. The reported mean or midpoint of each cumulative asbestos exposure category was log-transformed and meta-analyzed using multi-level, random intercept, log-linear regression models. Analyses explored higher risk groups and sources of heterogeneity by exposure and study characteristics, including the quality of the exposure assessment. The investigators reached consensus on interpretation of the findings by emphasizing the consistency of the findings and study informativeness, and placing less emphasis on statistical significance in the meta-analyses of a significant occupational hazard.Results The predicted exposure-response curve, extrapolated from the overall meta-risk estimate (theta (θ)=0.13 (95% CI, 0.10, 0.16)), was nonlinear with a steeper rise in risk at the lower exposure range: the estimated RR was 1.21 (1.17, 1.26) at 5 f-yr/ml; 1.43 (1.33, 1.53) at 25 f-yr/ml; 1.52 (1.40, 1.63) at 50 f-yr/ml; 1.61 (1.47, 1.64) at 100 f-yr/m; and 1.71 (1.54, 1.87) at 200 f-yr/ml. Stronger exposure-response relationships were observed in studies with higher quality exposure assessment (e.g., RR from 1.32 to 2.07 across the same range of exposures), as well as in studies of asbestos production and manufacturing workers, particularly in North America. The results were robust to sensitivity analyses investigating sources of heterogeneity and analytic methods.Conclusion Increasing cumulative asbestos exposure is associated with higher lung cancer risk across the range of cumulative exposure, with a steeper rise in risk at the lower exposure range.