Confounding of Ozone’s Cardiorespiratory Health Effects by Ozone Reaction Products: Evidence from Two Panel Studies with Contrasting Exposure Conditions

Personal ozone (O 3 ) exposure is often regarded as a more accurate metric of individual-level O 3 exposure relevant to health outcomes than outdoor O 3 concentration. However, epidemiological studies of personal O 3 exposure health effects have been inconsistent, including paradoxical “protective’ effects, suggesting the presence of confounding factors yet to be identified. When O 3 enters indoors, it reacts with indoor chemicals to form ozone reaction products (ORPs) that may adversely affect cardiorespiratory health. The correlation between personal ORPs and O 3 exposure can vary largely across individuals, which leads us to hypothesize ORPs exposure as a confounder for cardiorespiratory effects of personal O 3 exposure. We analyzed two complementary datasets from panel studies in 89 healthy adults and in 43 asthmatic children, each measured four times for cardiorespiratory biomarkers and personal exposures to O 3 , O 3 loss (a proxy for ORPs exposure), and PM 2.5 , respectively. We found personal O 3 and O 3 loss exposures were highly correlated in the Adults’ Study but poorly correlated in the Children’s Study. In the Adults’ Study, personal O 3 exposure was significantly associated with biomarkers of pulmonary oxidative stress, airway inflammation, lung function, systemic oxidative stress, thrombosis, arterial stiffness, and blood pressure. After adjusting for O 3 loss, the O 3 effects on many of these biomarkers were largely changed in terms of effect sizes and/or statistical significance. However, similar confounding effects were not observed in the Children’s Study. These findings may help explain the inconsistencies in the literature regarding the cardiorespiratory effects of personal O 3 exposure and highlight the need to account for potential confounding from ORPs in future studies.