To the Editor:
Recently, it was estimated that respirable elemental carbon exposure needs to be limited to below 1 μg/m3 to comply with the common regulatory action level of one additional lung cancer death per 1000 workers.1
Considerable doubts remain whether these calculations reflect the actual risk because of exposure to diesel exhaust. The calculations were based on a simulation study using the data from the Diesel Exhaust in Miners Study (DEMS).2 The authors treated respirable elemental carbon exposure, employment status, and employment location as time-varying variables. They treated sex, race, state, and an indicator for workers hired before dieselization as time-independent variables. However, the application of the G-formula is tied to some important requirements, especially the conditional exchangeability and the counterfactual consistency. As such, two more important confounders should have been taken into account, namely smoking habits and former exposures to occupational lung carcinogens. The investigators should have discussed the possibility that these confounders violate the requirements. Applying the counterfactual concept to the study, one may ask what would have happened if a miner had worked as a gardener, e.g., rather than as an underground miner. Although he would not have been exposed to diesel exhaust in this counterfactual case, his smoking habits could have been quite different, noting that smoking is prohibited in underground workplaces. Therefore, the consideration of smoking habits as a time-dependent variable is advisable. Likewise, former employment in other high-risk occupations needs to be taken into account. The DEMS revealed that the mortality due to other pneumoconiosis was markedly elevated for ever-underground workers (standardized mortality ratio =16.21 [95% confidence interval = 8.37, 28.32]; n = 12).2 It can be assumed that these 12 deaths are only the tip of the iceberg. The number of prevalent cases of pneumoconiosis in the cohort could even be higher than the number of lung cancer cases. Moreover, there is evidence that these diseases are progressive over time, even without further exposure to the causal agent.3 Therefore, workers with former employment in high-risk jobs should be completely excluded from the study base.
Finally, the “g-null paradox” needs to be considered. A recent review shows that the evidence for an exposure–response relationship between respirable elemental carbon and lung cancer is very limited.4 Given that the causal null hypothesis may be true, the estimated exposure limits are not reliable and, hence, should not be used for regulatory purposes.
Federal Institute for Occupational
Safety and Health
Division Work and Health
1. Neophytou AM, Picciotto S, Costello S, Eisen EA. Occupational diesel exposure, duration of employment, and lung cancer: an application of the parametric G-formula. Epidemiology. 2016;27:21–28.
2. Attfield MD, Schleiff PL, Lubin JH, et al. The Diesel Exhaust in Miners study: a cohort mortality study with emphasis on lung cancer. J Natl Cancer Inst. 2012;104:869–883.
3. Lee HS, Phoon WH, Ng TP. Radiological progression and its predictive risk factors in silicosis. Occup Environ Med. 2001;58:467–471.
4. Möhner M, Wendt A. A critical review of the relationship between occupational exposure to diesel emissions and lung cancer risk. Crit Rev Toxicol. 2017;47:185–224.