Sputum analysis, bronchial hyperresponsiveness, and airway function in asthma: Results of a factor analysis

Background: Recent studies have shown weak associations among FEV1, bronchi al hyperresponsiveness (BHR), sputum eosinophils, and sputum eosinophil cat ionic protein (ECP), suggesting that they are nonoverlapping quantities. Th e statistical method of factor analysis enables reduction of many parameter s that characterize the disease to a few independent factors, with each fac tor grouping associated parameters. Objective: The purpose of this study was to demonstrate, by using factor an alysis, that reversible airway obstruction, BHR, and eosinophilic inflammat ion of the bronchial tree, as assessed by cytologic and biochemical analysi s of sputum, may be considered separate dimensions that characterize chroni c bronchial asthma. Methods: Ninety-nine clinically stable patients with a previous diagnosis o f asthma underwent spirometry, sputum induction, and histamine inhalation t ests. Results: Most patients were nonobstructed (FEV1, 91% +/- 20%); a low level of bronchial reversibility (FEV1 increase after beta(2)-agonist, 7.8% +/- 9 .2%) and BHR (histamine PC20FEV1 geometric mean, 0.98 mg/mL) were found. Sp utum eosinophil differential count (12.4% +/- 17.7%) and sputum ECP (1305 /- 3072 mu g/mL) were in the normal range of our laboratory in 38 and 22 pa tients, respectively. Factor analysis selected 3 different factors, explain ing 74.8% of variability. Measurements of airway function and age loaded on factor I, PC20FEV1 and beta(2)-response loaded on factor II, and sputum EC P and eosinophils loaded on factor III. Additional post hoc factor analyses provided similar results when the sample was divided into 2 subgroups by r andomization, presence of airway obstruction, degree of BHR, percentage of sputum eosinophils, or concentration of sputum ECP. Conclusions: We conclude that airway function, baseline BHR, and airway inf lammation may be considered separate dimensions in the description of chron ic asthma. Such evidence supports the utility of routine measurement of all these dimensions.