Background: Indoor air pollution concentrations in households (HAP) span a wide range from those with open solid fuel combustion, averaging many hundreds of μg/m3 of PM2.5 over the year in poorly ventilated conditions, to relatively clean levels in homes using gas or electric heating and cooking devices. To assess the health risk remaining at each level above the WHO AQGs requires understanding the exposure-response relationships over far greater exposure ranges than are available from any one epidemiological study available today. As part of the GBD Project Comparative Risk Assessment (CRA), a set of integrated exposure-response (IER) functions were developed that link risks across more than three orders of magnitude in exposures from epidemiologic studies of combustion particle exposures from active and secondhand tobacco smoking, outdoor air pollution (OAP), and HAP. These are being considered for informing the development of the IAQGs. Methods: We briefly review the current state of development of the IERs for several important disease endpoints and their potential use in the IAQGs. Results: There have been three major uses of the IERs for informing policy. The main impetus for their development by the OAP CRA was to determine risks at exposure levels higher than available in the OAP epidemiologic studies. The main use by the HAP CRA was to interpolate risks for cardiovascular endpoints for which there is evidence at higher (active smoking) and lower (secondhand smoking/OAP) exposures, but no direct evidence yet for HAP. In addition, it was used to enable risk estimates in comparison to a common low counterfactual for both HAP and OAP. Conclusions: It has been proposed to use lERs to estimate the benefits of alternative interventions at different points along the exposure axis. We explore the potentially counterintuitive implications of doing so that arise because of the highly supralinear shape of most of the lERs and offer some ways forward.
展开▼
