Longitudinal personal environmental exposure and Streptococcus pneumoniae colonisation in a cohort of children 5-14 years old in rural Gambia.

Abstract Background Air pollution contributes substantially to morbidity and mortality in low-income countries, where reliance on solid fuels leads to high levels of household exposure. Colonisation of the upper respiratory tract by Streptococcus pneumoniae is a prerequisite for pneumococcal disease, and inhaled air pollutants may influence colonisation through impaired mucociliary clearance. Therefore, we investigated the association between S. pneumoniae colonisation and personal exposure to particulate matter (PM2.5) and carbon monoxide (CO) as well as temperature and relative humidity in older children in The Gambia. Methods We conducted a prospective study of 134 children aged 5-14 years across eight visits in rural Gambia. Personal 24-hour PM2.5, CO, temperature and relative humidity exposures were measured using personal monitors. Concurrently, combined nasopharyngeal/oropharyngeal swabs were collected at each visit, and S. pneumoniae colonisation was determined by microbiological culture. Mixed-effects logistic regression models were used to examine associations between environmental exposures and colonisation. Exposures were additionally decomposed into between-child and within-child components. Models were adjusted for age, season, and household income tertile. Results A total of 554 child visits with paired exposure and microbiological data were analysed. Overall, 24-hour geometric mean PM2.5 and CO were 12.6 {micro}g/m3 (95% CI 12.2-13.1) and 0.16 ppm (95% CI 0.14-0.19) respectively and overall S. pneumoniae colonisation prevalence was 39.7%. In conventional models, PM2.5, CO, temperature, and relative humidity were not significantly associated with S. pneumoniae colonisation after adjustment for confounders. In the within-between decomposition analysis, higher between-child temperature exposure was consistently associated with lower odds of colonisation (mutually adjusted model: OR 0.74, 95% CI 0.56-0.98; p = 0.034). Relative humidity showed an inverse within-child association in crude analysis (OR 0.70, 95% CI 0.52-0.93; p = 0.013), which attenuated after adjustment for season and other confounders. No associations were observed for PM2.5 or CO exposures. Conclusion At the exposure levels observed in this cohort of older children, PM2.5 and CO exposures were not independently associated with S. pneumoniae colonisation. The pattern observed for temperature suggests that longer-term temperature conditions may be more relevant to pneumococcal colonisation than short-term exposure variation in this setting.