Discriminating power of measurements from maximum expiratory flow-volume curves

Maximum expiratory flow-volume curves have been used to characterize normal and abnormal respiratory function. Previous investigators have usually attempted to study whether any single lung function separately is most useful in characterizing lung dysfunction. Using stepwise discriminant analysis we sought to identify which combination of lung function parameters derived from maximum expiratory flow-volume curves could best discriminate between epidemiologic groups defined on the basis of smoking status and presence or absence of respiratory symptoms. We studied 5,031 whites, 2,271 males and 2,760 females, from three communities. In general, combinations of lung functions were no more useful than single lung functions to best discriminate between the groups studied. Flow rates at 50 and 25% of vital capacity remaining to be expired were most often selected as the best discriminating lung function, possibly suggesting obstruction of small airways in the smoking and symptomatic groups.