Comments on “The burden of avoidable disease from air pollution: implications for prevention”

Samet and Buran in their paper published in your Journal recently [1] compare the burden of disease associated with air pollution to that of tobacco smoking. They suggest that, potentially, health gains due to elimination of smoking should be greater than those due to reduction of the exposure to air pollution. They also warn against an “artificial contest between tobacco control and air quality management”. I do agree with this warning and would like to emphasize several issues in support of investment in clean air policies as an important public health issue.
There are several aspects of the health burden estimates comparison in [1] which deserve a comment. The first relates to the estimated magnitude of the health burden attributable to air pollution. The numbers quoted by Samet and Buran come from the Global Burden of Disease (GBD) study and its results for 2017 [2]. One may notice, that this ambitious project publishes the results of the comparative risk assessment every year since 2010. The improvement of methods and input data results in changes of the estimates produced in subsequent releases of this, and other, analyses [3]. The estimates of the burden of disease due to air pollution differ also between the assessments made by GBD project, WHO [4], EEA [5] or other authors [6] for Poland. Here the corresponding estimates of the annual number of deaths attributed to air pollution are, respectively (in thousands): 23, 27, 44 and 58. Big differences between these results may be confusing if details of the analysis are not communicated or understood. They depend on the health outcomes considered, concentration-response functions used, counterfactual level of exposure providing a point of reference for the assessments as well as the data on population exposure.
While GBD and WHO analyses are based on the relation of few, strictly defined health outcomes (e.g. ischemic heart disease, cerebrovascular disease, lung cancer and lower respiratory infections) to the exposure, the EEA analysis considers deaths from all causes, and Liliveld et al.: all non-communicable diseases and respiratory infections. Possible under-reporting of the specific causes of death in the Polish mortality data (with unproportionate number of IHD or stroke deaths registered as other diseases, such as atherosclerosis) might be the reason of under-estimation of the effects by GBD and WHO assessments based on strictly selected causes of death [7, 8]. Widening of the...


View full text...