Global Burden of Disease and Air Pollution

How do you count deaths from something nobody dies of directly? Air pollution doesn't appear on death certificates. People die of heart attacks, strokes, lung cancer, pneumonia, COPD. The Global Burden of Disease (GBD) study, run by the Institute for Health Metrics and Evaluation at the University of Washington, has spent two decades building the statistical machinery to estimate how many of those deaths are pushed forward in time by dirty air.

Background

GBD began in the early 1990s as a World Bank and WHO project led by Christopher Murray and Alan Lopez. IHME now updates it roughly every two years with hundreds of collaborators. The State of Global Air report, a joint product of IHME and the Health Effects Institute (HEI) in Boston, translates the GBD air pollution estimates into a public-facing annual update.

How the estimates are built

The method is called comparative risk assessment. It runs in five steps:

Estimate population exposure to a pollutant by combining ground monitors, satellite retrievals (mostly from NASA's MODIS and the Aura mission), and chemical transport models on roughly a one-kilometer grid.
Build an exposure-response curve from cohort studies that links pollution levels to relative risk for each disease.
Calculate the population attributable fraction for each disease in each country.
Multiply by total disease deaths to get attributable deaths.
Convert deaths and illness into disability-adjusted life years (DALYs).

The PM2.5 exposure-response curve, called the Global Exposure Mortality Model (GEMM), was developed by Richard Burnett and colleagues and published in PNAS in 2018. It draws on 41 cohort studies covering low, middle, and high exposure ranges, including Chinese male smokers to extend the curve into very polluted air.

What the latest numbers say

GBD 2021, summarized in the State of Global Air 2024:

Air pollution contributed to 8.1 million deaths worldwide in 2021, making it the second-leading global risk factor for death after high blood pressure.
Ambient PM2.5 alone accounted for roughly 7.8 million of those deaths.
Household air pollution (mostly indoor cooking with solid fuels) caused about 3.1 million deaths and led to roughly 500,000 child deaths under age five, concentrated in sub-Saharan Africa and South Asia.
Ozone contributed to about 489,000 deaths, mostly from COPD.
The leading diseases driven by ambient PM2.5 were ischemic heart disease, stroke, COPD, lung cancer, lower respiratory infections, type 2 diabetes, and adverse birth outcomes (a category added in the 2019 update after evidence on low birth weight and preterm birth firmed up).

Why it matters

Putting a number on air pollution makes it comparable to other risks. GBD lets a health minister see that PM2.5 in their country kills more people than road injuries, HIV, or malaria. It also lets researchers track progress over time: ambient PM2.5 deaths in China appear to have peaked around 2015 and are now declining, while South Asia's burden continues to rise.

The estimates also feed into cost-benefit analyses for regulation. The EPA's regulatory impact analyses for the 2024 PM2.5 NAAQS revision cited GBD-derived numbers, as do the EU's air quality directive impact assessments.

Open questions

Uncertainty in the exposure-response curve is largest at the high and low ends. The shape of the curve below about 10 micrograms per cubic meter still relies on a smaller number of cohorts, mostly in North America and Europe. Adding new diseases (dementia, mental health outcomes, kidney disease) is an active area of research and may push the global number higher in future updates.

Sources

This article is for educational purposes only. Canairy does not provide medical advice, diagnosis, or treatment. Talk to a qualified health professional about your specific situation.