Time domain and flow indices of bronchial hyperresponsiveness: association with asthma symptoms, atopy and smoking.

Conventional measures of bronchial hyperresponsiveness (BHR) are only weakly associated with respiratory symptoms in epidemiological studies. Partial and maximal forced expiratory manoeuvres were recorded during histamine challenge testing in 1,959 young male farmers. Analysis was performed to test whether novel measures of BHR, using alternative flow and time domain indices, are more closely associated with asthma symptoms, smoking status and atopy than forced expiratory volume in one second (FEV1) and conventional measures of BHR. The first moments to 75% and 90% of the forced vital capacity (FVC) were calculated from full (F) and partial (P) forced expiratory manoeuvres (i.e. alpha1 75%F, alpha1 75%P), together with the instantaneous flows when 40% and 30% of the FVC remained to be expired (MEF40 and MEF30). BHR was measured by the provocative dose causing a 20% change (PD20) in the FEV1 and alpha1 75%, and also by the method of log dose slopes (LDS). Asthma was diagnosed from symptoms associated with asthma in 158 (8.1%) of the subjects. PD20 FEV1 could only be recorded in 190 subjects (9.7%), of whom only 48 had asthma, whereas LDSFEV1 was recorded in 1,725 (88%) subjects. From the prechallenge data, alpha1 75%, expressed as standardised residuals, showed the largest difference between smokers with and without asthma symptoms, and no indices showed significant differences between nonsmokers with and without asthma symptoms. From BHR data in both smokers and nonsmokers, LDSFEV1 showed one of the largest differences between those with and without asthma symptoms. With smoking status and atopy accounted for, the greatest partial correlation with asthma diagnosis was found for LDSMEF40P, and then for LDSFEV1, but LDSMEF40P was measurable in only just over one-half of the subjects. The authors conclude that time-domain indices are promising measures for longitudinal epidemiological studies concerning the relationship between bronchial hyperresonsiveness and environmental exposures. However, indices from the partial flow-volume loop suffer from censored data.