Poster abstract for HEI Annual Conference 2022
Integrating Satellites, Ground Monitoring, and Modeling to Estimate Long-Term NO2 Exposures and Associated Pediatric Asthma Impacts
Susan C. Anenberg1, Michael Brauer2,3, Katrin Burkart2, Dan Goldberg1, Perry Hystad4, Gaige Kerr1, Andrew Larkin4, Sarah Wozniak2
1George Washington University, Washington, DC, USA; 2Institute for Health Metrics and Evaluation, University of Washington, Seattle, WA, USA; 3University of British Columbia, Vancouver, BC, Canada; 4Oregon State University, Corvallis, OR, USA
Background. Recent meta-analyses of epidemiological studies from North America, Latin America, Europe, and East Asia show that traffic-related ambient NO2 pollution is associated with pediatric asthma incidence. A previous disease burden assessment found that NO2 could be responsible for 13% (~4 million) of new pediatric asthma cases globally, and up to nearly 50% in cities, but neither NO2 nor air pollution impacts on asthma were included in the Global Burden of Disease Study (GBD) prior to the GBD 2020 Study. Our objective is to estimate surface NO2 concentrations and associated pediatric asthma burdens from 1990-2020 globally for incorporation into the GBD 2020 Study and beyond.
Methods. We develop methods for the GBD 2020 Study by: 1) assessing the strength of the epidemiologic evidence linking NO2 and pediatric asthma incidence using a meta-regression tool used widely across GBD risk factor teams; 2) scaling an existing global land use regression (LUR) model for NO2 from 2010-2012 to 1990-2020 using satellite and atmospheric model data; 3) estimating the population attributable fraction (PAF) of NO2 on pediatric asthma incidence globally from 1990-2020. We also develop a new global LUR model using satellite, land use, and ground monitor data to predict daily, monthly, and annual NO2 concentrations at 50m resolution globally from 2005-2020 for use in future GBD iterations.
Results. Our NO2-pediatric asthma meta-analysis included 27 studies and found a relative risk of 1.082 (95% uncertainty interval: 1.010 – 1.160) at 5 ppb annual average NO2 concentration relative to 0 ppb, translating to a beta coefficient of 0.0155 (95% UI: 0.0085 – 0.0224) per 1 ppb NO2 concentration. Scaling the 2010-2012 LUR to other years using satellite and model data, we estimated that the population-weighted average annual mean NO2 concentration was 6.2 ppb in 2020, down from 7.5 in 1990. We estimated that PAFs dropped from 1990 to 2020 globally, with regional differences. Preliminary results indicate that PAFs dropped in high-income countries, Latin America/Caribbean, central and eastern Europe; central, southeast, and east Asia; and Oceania. PAFs rose in south Asia, sub-Saharan Africa, and north Africa and the Middle East. Finally, our new global daily LUR model for NO2 at 50m resolution has 6% greater global adjusted R2 (0.47 for daily predictions, 0.59 for monthly, and 0.63 for annual) than the previous 2010-2012 LUR model while concomitantly improving spatial and temporal resolutions.
Conclusions. Combining state of the science epidemiological and atmospheric information, we found that ambient NO2 contributes substantially to pediatric asthma incidence globally, particularly in cities. Mitigating traffic-related air pollution should be a crucial element of public health strategies for children.