Air pollution is associated with a distinct pathogen spectrum in community-acquired pneumonia: a prospective multicentre cohort study

Background Community-acquired pneumonia (CAP) is the leading cause of death due to infection worldwide, with its incidence and disease course influenced by air pollution exposure. Little is known about how air pollution influences microbiological findings in patients with CAP. Methods Patients were recruited from a prospective, multicentre study of CAP (CAPNETZ). Air pollution data were collected at the city level considering data up to one month before inclusion. Associations with sputum pathogen detection and cell counts were tested using generalised linear mixed-effects logistic regression models adjusting for clinical confounders, centre-level heterogeneity and meteorological data. Internal validation was performed using an Ensemble Feature Selection (EFS) approach. Results Among 4,446 eligible patients with available sputum samples, a potential pathogen was detected in 1,581 patients, while complete sputum-cell-count data were available for 4,014 patients. Detection of Streptococcus pneumoniae was associated with previous higher exposure to sulphur dioxide (SO 2 ) (OR 1.75, p = 0.008) and carbon monoxide (CO) (OR 1.57, p = 0.02). Escherichia coli was more frequently detected in patients who were exposed to elevated levels of nitrogen dioxide (NO 2 ) (OR 2.75, p = 0.004) and higher temperature (OR 1.51, p = 0.03), whereas increased fine particulate matter (PM 2.5 ) exposure (OR 0.37, p = 0.001) was linked to reduced detection. Furthermore, SO 2 exposure influenced the likelihood of obtaining a diagnostically relevant sputum culture and was associated with higher odds of detecting a high leucocyte count (OR 1.18, p = 0.02). Results were largely corroborated by EFS, particularly for S. pneumoniae and E. coli detection. Conclusion Exposure to air pollution was associated with a distinct pathogen spectrum in patients with CAP. These findings may have implications for the interpretation of microbiological data in an epidemiological and clinical context and highlight the importance of considering air pollution exposure as a confounding factor. Further validation in cohorts with high-quality environmental, microbiological and antimicrobial resistance data is needed. Study registration Registry: German Clinical Trials Register (DRKS), DRKS-ID: DRKS00005274, Registration date: 2013-09-11.