Abstract
Nonspecific bronchial hyperresponsiveness (BHR) is a hallmark of clinical asthma, but can be present in nonasthmatics as well. The diagnosis of asthma is based on clinical grounds, and no laboratory procedure can definitely establish its presence. This poses a problem in studies of asthma. If epidemiological studies are to provide valid information, the tools used must have a relative degree of predictive or diagnostic ability. This report determined whether the American Thoracic Society-Division of Lung Disease (ATS-DLD) respiratory questionnaire has the ability to predict different degrees of non-specific BHR. In the years 1983-1990, when the ATS-DLD questionnaire was used in our Natural History of Asthma study, 192 subjects completed the ATS-DLD questionnaire and underwent a standardized methacholine challenge. A recursive partitioning analysis of the ATS-DLD questionnaire was able to predict which questions would likely be angered if the subject had nonspecific bronchial reactivity to inhaled methacholine of 100 and 200 breath units. Positive responses for questions concerning treatment for asthma, wheezing, or shortness of breath, and emergency treatment for asthma predicted the presence of increased bronchial reactivity.