Dose-response-relationship between occupational exposure to diesel engine emissions and lung cancer risk: A systematic review and meta-analysis

Background: In 2012, the International Agency for Research on Cancer (IARC) concluded that diesel engine emissions (DEE) emissions cause cancer in humans. However, there is still controversy surrounding this conclusion, due to several studies since the IARC decision citing a lack of evidence of a dose -response relationship. Objectives: Through a systematic review, we aimed to evaluate all evidence on the association between occupational DEE and lung cancer to investigate whether there is an increased risk of lung cancer for workers exposed to DEE and if so, to describe the dose -response relationship. Methods: We registered the review protocol with PROSPERO and searched for observational studies in relevant literature databases. Two independent reviewers screened the studies' titles/abstracts and full texts, and extracted and assessed their quality. Studies with no direct DEE measurement but with information on length of exposure for high -risk occupations were assigned exposure values based on the DEE Job -Exposure -Matrix (DEEJEM). After assessing quality and informativeness, we selected appropriate studies for the dose -response metaanalysis. Results: Sixty-five reports (from thirty-seven studies) were included in the review; one had a low risk of bias (RoB) (RR per 10 mu g/m3-years: 1.014 [95%CI 1.007-1.021]). There was an increased, statistically significant risk of lung cancer with increasing DEE exposure for all studies (RR per 10 mu g/m3-years = 1.013 [95%CI 1.004-1.021]) as well as for studies with a low RoB in the exposure category (RR per 10 mu g/m3-years = 1.008 [95% CI1.001-1.015]). We obtained a doubling dose of 555 mu g/m3-years for all studies and 880 mu g/m3-years for studies with high quality in the exposure assessment. Discussion: We found a linear positive dose -response relationship for studies with high quality in the exposure domain, even though all studies had an overall high risk of bias. Current threshold levels for DEE exposure at the workplace should be reconsidered.