Which decreases in air pollution should be targeted to bring health and economic benefits and improve environmental justice?

Background: Fine particulate matter (PM2.5) exposure entails large health effects in many urban areas. Public measures aiming at decreasing air pollution are often designed without targeting an explicit health benefit. Our objective was to investigate the health and economic benefits and the social inequalities in exposure resulting from several scenarios of reduction of PM2.5 exposure, in order to support decisions about urban policies. Material and methods: In the French conurbations of Grenoble and Lyon (0.4 and 1.4 million inhabitants, respectively), PM2.5 yearly average exposure was estimated on a 10-m grid by coupling a PM2.5 dispersion model to population density. Changes in death cases, life expectancy, lung cancer and term low birth weight incident cases as well as associated health economic costs were estimated for ten PM2.5 reduction scenarios differing in terms of amplitude of reduction and spatial extent. Changes in social differences in PM2.5 exposure were also assessed. Results: During the 2015–2017 period, PM2.5 average exposure was 13.9 μg/m3 in Grenoble and 15.3 μg/m3 in Lyon conurbations. Exposure to PM2.5 led to an estimated 145 (95% Confidence Interval, CI, 90–199) and 531 (95% CI, 330–729) premature deaths, 16 (95% CI, 8–24) and 65 (95% CI, 30–96) incident lung cancers, and 49 (95% CI, 19–76) and 193 (95% CI, 76–295) term low birth weight cases each year in Grenoble and Lyon conurbations, respectively, compared to a situation without PM2.5 anthropogenic sources, i.e. a PM2.5 concentration of 4.9 μg/m3. The associated costs amounted to 495 (Grenoble) and 1767 (Lyon) M€/year for the intangible costs related to all-cause non-accidental mortality and 27 and 105 M€ for the tangible and intangible costs induced by lung cancer. A PM2.5 exposure reduction down to the WHO air quality guideline (10 μg/m3) would reduce anthropogenic PM2.5-attributable mortality by half while decreases by 2.9 μg/m3 (Grenoble) and 3.3 μg/m3 (Lyon) were required to reduce it by a third. Scenarios focusing only on the most exposed areas had little overall impact. Scenarios seeking to reach a homogeneous exposure in the whole study area were the most efficient in alleviating social inequalities in exposure. Conclusions: Reduction scenarios targeting only air pollution hotspots had little expected impact on population health. We provided estimates of the PM2.5 change required to reduce PM2.5-attributable mortality by one third or more. Our approach can help targeting air pollution reduction scenarios expected to entail significant benefits, and it could easily be transposed to other urban areas. Keywords: Dispersion model, Economic costs, Health impact assessment, Lung cancer, Fine particulate matter (PM2.5), Social deprivation