Abstract
Climate change has been assumed as a key contributor to global undernutrition. One possible channel is that global warming increases air pollution exposure by causing more and more wildfires. This study examined the effect of wildfire-sourced particulate matter (PM)(2.5) exposure on anaemia in women of reproductive age (WRA), a dominating form of undernutrition across the low- and middle-income countries (LMICs). We collected haemoglobin measurements of 1 549 972 WRA from 85 Demographic and Health Surveys (DHS) in 45 LMICs between 2004 and 2021. Wildfire-sourced PM(2.5) exposure was estimated through a synthesis of ground observations, satellite remote sensing measurements and the Goddard Earth Observing System with Chemistry (GEOS-Chem) model simulations. The effects on the haemoglobin levels, as well as the odds of anaemia, were estimated by the mixed-effects regression models. Finally, population-attributable fraction or number was calculated to assess the burden of wildfire-sourced anaemia among WRA, by extrapolating the regression model result to a study domain of 125 LMICs. In the fully adjusted model, a 1 μg/m³ increase in wildfire-sourced PM(2.5) was associated with a 0.1263 g/L reduction in the haemoglobin levels of women (95% CI -0.1540 to -0.0987) and a 1.3% increase in odds of anaemia (ORs 1.013, 95% CI 1.010 to 1.016). The non-linear analysis showed a monotonically increasing exposure-response relationship, which suggested a saturated effect for an exposure concentration >5.85 μg/m³. Across the study domain of 125 LMICs, wildfire-sourced PM(2.5) contributed to WRA anaemia of 35.9 million (95% CI 30.1 to 42.1) in 2000, rising to 52.8 (95% CI 44.3 to 61.9) in 2019. The numbers accounted for 8.81% (95% CI 7.38% to 10.32%) and 9.54% (95% CI 7.99% to 11.17%) of all anaemic WRA, respectively. To meet the global 2030 target of reducing 50% WRA anaemia, warranted are climate mitigations and adaptations against wildfire smoke exposure.