Premature mortality and economic losses attributable to landscape fire-related PM2·5 in ten countries and territories: a causal analysis.

BACKGROUND

Landscape fire-related fine particulate matter (LF-PM2·5) is associated with premature mortality. However, the mortality-related economic burden of LF-PM2·5 remains poorly quantified. We aimed to estimate the short-term effects of LF-PM2·5 on mortality and the associated economic losses.

METHODS

In this causal analysis, we linked daily mortality records from 1366 communities in ten countries and territories during 2000-19 to population-weighted LF-PM2·5 concentrations estimated from daily modelled data. Economic losses were estimated using adjusted values of a statistical life. We estimated community-specific associations using instrumental variable models and pooled them using random-effects meta-analysis. We compared instrumental variable estimates with traditional association estimates as well as corresponding estimates for non-LF-PM2·5.

FINDINGS: We analysed 47 844 851 deaths, corresponding to mortality-related economic losses valued at US$60·6 trillion. Each 1 μg/m3 increase in the 2-day moving average of LF-PM2·5 was associated with a 0·60% (95% CI 0·19-1·00) increase in mortality and a 0·59% (0·21-0·97) increase in economic losses. An estimated 906 869 deaths and $994·5 billion in economic losses were attributable to LF-PM2·5, with attributable fractions of 1·90% for deaths and 1·64% for economic losses. Traditional association estimates were found to be lower than instrumental variable estimates by a median of 56·0% (2·2-111·3) for mortality and 62·3% (9·0-116·1) for economic losses. LF-PM2·5 represented 19·0% of ambient PM2·5 mass but accounted for 30·6% of PM2·5-attributable deaths and 27·7% of PM2·5-attributable economic losses.

INTERPRETATION: Short-term exposure to LF-PM2·5 was associated with substantial mortality and economic losses across diverse settings. The disproportionate contribution of LF-PM2·5 to PM2·5-attributable burden, together with the larger instrumental-variable estimates, suggests that traditional assessments might understate the consequences of landscape fire smoke.

FUNDING: National Health and Medical Research Council, China Scholarship Council.