Impact of the Ultra Low Emission Zone on lung function growth in children in London, UK: a prospective parallel cohort study

Background Traffic-related air pollution is a risk factor for lung disease and early mortality. Clean air zones are public health policy interventions used to reduce traffic-related air pollution in urban areas, but evidence of their health benefits is limited. We describe a natural experiment study evaluating the impact of the Ultra Low Emission Zone (ULEZ) in London, UK. The primary aim was to assess the impact of the ULEZ on children's lung function growth trajectories, by comparing forced expiratory volume in 1 s (FEV 1 ) measurements over 5 years between children in London and Luton, UK. Methods The Children's Health in London and Luton (CHILL) study is a prospective, two-arm, parallel cohort study. We recruited children aged 6–9 years from primary schools in central London (the original area of the ULEZ implementation) and Luton (a comparator site with no clean air zone). Children were excluded if they had symptoms of lung disease (excluding asthma) or learning or physical disabilities preventing them from giving informed assent. Lung function was measured by spirometry at annual school visits at baseline (before ULEZ implementation) and over the following 4 years. Annual residential exposures to nitrogen dioxide (NO 2 ) and particulate matter with aerodynamic diameters of less than 10 μm (PM 10 ) and less than 2·5 μm (PM 2·5 ) were estimated at each child's home address at 20 m 2 resolution using a validated dispersion modelling system. The primary outcome was the annual rate of lung function growth, measured as post-bronchodilator FEV 1 over the 5-year study period. We assessed growth trajectories against individualised residential exposure to NO 2 , PM 10 , and PM 2·5 . We used mixed-effects linear regression to compare lung function growth between London and Luton and to examine associations between air pollution exposures and lung growth. Findings Of 122 schools approached, 84 (69%) agreed to participate (44 schools in London; 40 schools in Luton). Of 9419 children invited, we recruited 3414 children (1664 in London and 1750 in Luton) between June 5, 2018, and April 4, 2019, before ULEZ implementation. 3209 (94·0%) of 3414 children were included in the FEV 1 analysis (1557 children in London, 1652 children in Luton). Children were similar across the two sites in terms of age, height, and weight, with a higher proportion of girls in London (867 [56%] of 1557) than in Luton (809 [49%] of 1652). Baseline adjusted FEV 1 was lower in London than Luton (difference –38 mL, 95% CI –58 to –18; p=0·0002). At baseline, children's modelled annual exposures to NO 2 , the pollutant most reflective of exhaust emissions, were 18·95 μg/m 3 (95% CI 18·51 to 19·38; p