Abstract
Tests of nonspecific airway hyperresponsiveness are frequently used in the study of asthma both in the clinical settings and in epidemiologic studies. However, standard methods for characterizing individual tests and comparing results between subgroups have not been established. The most frequently used method of characterizing response curves is to report the dose that results in a 20% fall in FEV1 from the initial or baseline FEV1 value (PD20FEV1). Other investigators have suggested using the response slopes. In this study we demonstrate an alternative method of analysis that uses all of each subject's response data, makes comparisons between subgroups, and can include explanatory covariables. This approach is demonstrated using methacholine challenge data obtained in New Zealand children at 9 and 11 yr of age. The results showed significant differences between the mean dose-response curves of wheezers and nonwheezers, that responsiveness increased with the frequency of reported wheeze, and that initial pulmonary function and serum IgE are significantly related to responsiveness. These factors were not always significant using more traditional methods of analysis, indicating an increased sensitivity with this method of analysis.