Mixture analysis identifies ozone as driving factor for child asthma symptoms.

Background: The natural distributions of ambient air pollutants are often correlated. Existing studies have found that exposures to various air pollutants are associated with elevated risks of asthma symptoms among children. However, most studies applied single-pollutant models, which cannot distinguish between causal effects and associations due to correlations with other measured or unmeasured pollutants. Objective: We sought to investigate air pollutant mixtures and child asthma symptoms and identify key risk factors. Methods: The Children's Health and Air Pollution Study recruited 299 children in Fresno, California, 63 of whom had ever-diagnosed asthma. We assessed the children's prior 12-month exposures to 8 ambient air pollutants, namely, particulate matter with aerodynamic diameter of 2.5 μm, particulate matter with aerodynamic diameter of 10 μm, nitrogen oxides, nitrogen dioxide, ozone, carbon monoxide, elemental carbon, and polycyclic aromatic hydrocarbons, and asthmatic symptoms (wheeze and cough) at 2 visits, at age approximately 7 and 9 years. We conducted repeated-measures analysis with mixture analysis methods, including principal-component analysis and quantile-based g-computation (q-gcomp). Results: The 8 air pollutants exhibited strong intercorrelation. In single-pollutant models, exposure to ozone was associated with higher risk of cough (odds ratio, 1.39; 95% CI, 1.06-1.82). Using principal-component analysis and q-gcomp, exposures to nitrogen oxides, elemental carbon, and ozone had relatively high contributions to cough and wheeze. The association between ozone and cough was consistently positive from single-pollutant models, double-pollutant models, principal-component analysis, and quantile-based g-computation with negative control. Conclusion: Ozone stands out among the 8 air pollutants and may be a driving risk factor for persistent cough among children with asthma.