Key research papers on air pollution and health

How the literature is organised

The field is built on epidemiology. Dockery et al. (1993) estimated the effect of air pollution on mortality with cohort data from six US cities, controlling for individual risk factors, and time-series studies such as Schwartz (1999) on hospital admissions for heart disease in eight US counties showed short-term effects. Pope and Dockery (2006) drew the lines of evidence together in a long critical review, and Brunekreef and Holgate (2002) did the same for a clinical readership in The Lancet.

Mechanism and outcome-specific work followed. Brook et al. (2010) reviewed the evidence on cardiovascular disease for the American Heart Association and concluded that the relationship with fine particles is causal. Kampa and Castanas (2008) review the human health effects of air pollution in general, and Kim, Kabir and Kabir (2015) those of airborne particulate matter.

A third group quantifies burden. Lelieveld et al. (2015) apportioned premature mortality to source categories, and Cohen et al. (2017) reported 25-year trends in the burden of disease attributable to ambient air pollution.

Main debates

The American Heart Association writing group judged the evidence consistent with a causal relationship between fine particles and cardiovascular morbidity and mortality (Brook et al., 2010). The open questions in the listed papers are of a different kind. One is whether associations persist below current limit values, which Strak et al. (2021) and Stafoggia et al. (2022) were designed to test. Another is whether particles from different sources are equally toxic: Lelieveld et al. (2015) assume that they are and then vary the assumption in a sensitivity analysis. A third is how to separate correlated pollutants, which Schwartz (1999) addressed by comparing counties where particles, carbon monoxide and weather were correlated to different degrees. Burden estimates depend on the exposure-response functions chosen (Cohen et al., 2017).

Where recent work is heading

Large cohorts now test low concentrations (Stafoggia et al., 2022; Strak et al., 2021). Health impact assessments for individual cities (Khomenko et al., 2021; Southerland et al., 2022) and source apportionment at global scale (McDuffie et al., 2021) bring the evidence closer to policy. Fisher et al. (2021) examine air pollution and development in Africa, where household air pollution is still the predominant form but is declining while ambient pollution increases.