Analysis of dose-response curves to methacholine. An approach suitable for population studies

Abstract

Dose-response curves to methacholine were examined in 9 normal and 10 asthmatic volunteers to determine whether the relationship between dose and response can be adequately summarized by means of a single, continuous measure that is not censored at lower levels of bronchial responsiveness. Subjects underwent a standard methacholine challenge test. There was a strong linear relationship between percent decline FEV1 and cumulative dose methacholine. We summarized each dose-response curve by the slope of a line extending from the origin to the last data point obtained. This summary dose-response slope effectively separated asthmatic from normal subjects, and there was a greater than 3,000-fold difference between the least and most responsive subjects. There was a high degree of correlation between the dose-response slope determined by the standard methacholine challenge protocol and that determined by an abbreviated protocol currently being used to examine nonspecific airway responsiveness in a large, longitudinal study of aging. Among the participants of the latter study, there is a unimodal, skewed distribution of dose-response slope. Dose-response slope is proposed as a quantitative measure of nonspecific airway responsiveness that avoids censoring and that may be particularly useful in epidemiologic studies.