Measurement of bronchial responsiveness to hypertonic saline was applied in 22 study centers worldwide as part of Phase Two of the International Study of Asthma and Allergies in Childhood (ISAAC Phase Two). Because the amount of inhaled saline was difficult to standardize during the stepwise protocol with inhalation periods of increasing duration, we evaluated different statistical procedures based on inhalation time in relation to wheeze and current asthma. Data on random samples on 9 to 11-year-old children (n = 1,418) from two German centers were analyzed. The following statistical approaches were evaluated: (1) bronchial hyperreactivity (BHR) defined dichotomously as a fall in FEV1 (forced expiratory volume in 1 s) >or=15%; (2) PT15: the provocation time causing BHR using survival-analyses methods; (3) time-response-slope (continuous) of the individual FEV1-courses calculated by a linear model after comparing different mathematical models. The sensitivity and specificity of BHR versus current asthma were 47% and 87%, respectively. Analyses of the provocation time indicated an increased risk (adjusted hazard-ratio: 4.3; 95% CI: 2.8-6.5) for a fall in FEV1 >or= 15% among children with current asthma in comparison to those without. The time-response-slope differed markedly between children with and without wheeze and current asthma (P < 0.0001). BHR is meaningful and relatively easy to use, but has low sensitivity. Time-response-slopes utilize the available information from the stepwise protocol better than BHR and survival-analysis based on PT15. Response parameters based on inhalation time discriminate well between children with and without asthma and will be compared in the analyses of ISAAC Phase Two data.
(c) 2006 Wiley-Liss, Inc.