Predicting respiratory morbidity from pulmonary function tests: a reanalysis of ozone chamber studies.

Some consequences of acute exposure to ozone are best measured in studies of human respiratory responses in controlled exposure chambers. These studies typically examine relationships between exposures to alternative pollutant concentrations and indicators of lung function as measured by spirometry, such as forced expiratory volume in one second, FEV1. However, the association of respiratory morbidity with these changes in lung function is not well established. To gain a better understanding of the relationship between ozone-related changes in pulmonary function and respiratory symptoms, data from several clinical studies have been reanalyzed. Logistic regression models were used to determine the quantitative relationship between changes in FEV1 and the probability of a mild or moderate lower respiratory symptom. Models were developed that corrected for repeated sampling of individuals and both population-averaged and subject-specific effects were determined. The results indicate the existence of a strong and consistent quantitative relationship between changes in lung function and the probability of a respiratory symptom. Specifically, a 10 percent reduction in FEV1 is associated with a 15 percentage point increase in the probability of a mild, moderate or severe lower respiratory symptom and a 6 percentage point increase in the probability of a moderate or severe lower respiratory symptom.