A chronology of ratios between black smoke and PM(10) and PM(2.5) in the context of comparison of air pollution epidemiology concentration-response functions

BACKGROUND: For many air pollution epidemiological studies in Europe, ‘black smoke’ (BS) was the only measurement available to quantify ambient particulate matter (PM), particularly for exposures prior to the mid-1990s when quantification via the PM(10) and/or PM(2.5) metrics was introduced. The aim of this work was to review historic BS and PM measurements to allow comparison of health concentration-response functions (CRF) derived using BS as the measure of exposure with CRFs derived using PM(10) or PM(2.5). METHODS: The literature was searched for quantitative information on measured ratios of BS:PM(10), BS:PM(2.5), and chemical composition of PM; with specific focus on the United Kingdom (UK) between 1970 and the early 2000s when BS measurements were discontinued. RESULTS: The average BS:PM(10) ratio in urban background air was just below unity at the start of the 1970s, decreased rapidly to ≈ 0.7 in the mid-1970s and to ≈ 0.5 at the end of the 1970s, with continued smaller declines in the 1980s, and was within the range 0.2–0.4 by the end of the 1990s. The limited data for the BS:PM(2.5) ratio suggest it equalled or exceeded unity at the start of the 1970s, declined to ≈ 0.7 by the end of the 1970s, with slower decline thereafter to a range 0.4–0.65 by the end of the 1990s. For an epidemiological study that presents a CRF (BS) value, the corresponding CRF (PM10) value can be estimated as R (BS:PM10) × CRF (BS) where R (BS:PM10) is the BS:PM(10) concentration ratio, if the toxicity of PM(10) is assumed due only to the component quantified by a BS measurement. In the general case of some (but unknown) contribution of toxicity from non-BS components of PM(10) then CRF (PM10) > R (BS:PM10) × CRF (BS), with CRF (PM10) exceeding CRF(BS) if the toxicity of the other components in PM(10) is greater than the toxicity of the component to which the BS metric is sensitive. Similar analyses were applied to relationships between CRF (PM2.5) and CRF (BS). CONCLUSIONS: Application of this analysis to example published CRF (BS) values for short and long-term health effects of PM suggest health effects from other components in the PM mixture in addition to the fine black particles characterised by BS. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s12940-017-0252-2) contains supplementary material, which is available to authorized users.