Related Experiment Video
Updated: Aug 31, 2026
Real-World M3-BREATHE: Toward Multimodal Mobile Monitoring of Behaviour, Respiration, and Exposures for Treatment and Health Evaluation
Published on: June 5, 2026
Air Pollution Impairs Marathon Performance: A Cross-Sectional Analysis of 2.7 Million Finishers Across Six World
James A Donaldson1,2,3, Samuel Cai4,5,6,7, Josh Vande Hey4,5,6,7
1Division of Cardiovascular Sciences, University of Leicester, Leicester, UK. jad72@leicester.ac.uk.
Background:
Ambient air pollution impairs respiratory and cardiovascular function during exercise, yet evidence on its effects across the full spectrum of endurance performance remains limited.
Objectives:
To quantify the association between race-day ambient air pollution and marathon finishing time across six World Marathon Majors, and to determine whether effects vary by sex, age group, and performance level.
Methods:
We analysed 2,756,553 net finishing times from six World Marathon Majors (Berlin, Boston, Chicago, London, New York City, and Tokyo) for editions held between 2010 and 2024. Race-day nitrogen dioxide (NO₂) and fine particulate matter (PM₂.₅) concentrations were obtained from the Copernicus Atmosphere Monitoring Service (CAMS) global reanalysis dataset. Linear mixed-effects models with random intercepts for each of 75 marathon-year events estimated pollution effects on finishing time, adjusting for heat index, wind speed, and temporal trends. Stratified analyses were conducted among 1,815,188 runners with available age data.
Results:
Of the pollutants examined, NO₂ showed the stronger association with finishing time. Each 1 µg/m3 increase was associated with 1.43 min slower finishing time in men (95% confidence interval [CI] 0.66 to 2.21) and 1.56 min in women (95% CI 0.63 to 2.48). PM₂.₅ effects were not statistically significant: -0.13 min per µg/m3 in men (95% CI -0.67 to 0.40) and -0.14 min per µg/m3 in women (95% CI -0.76 to 0.49). Recreational runners experienced NO₂ effects approximately six-fold larger than elite runners (2.12 vs 0.34 min per µg/m3); expressed as a percentage of mean finishing time, this gradient persisted (0.67% vs 0.18%). Younger runners (18-29 years) showed larger effects than older runners (2.17 vs 1.32 min per µg/m3 for ≥60 years). Kolmogorov-Smirnov tests confirmed that higher NO₂ was associated with rightward shifts across the entire finishing time distribution.
Conclusions:
Ambient air pollution, particularly NO₂, is associated with impaired marathon performance in a dose-dependent manner, with the largest effects among recreational runners. These findings have implications for race management and public health policy regarding outdoor physical activity in urban environments.
