[en] BACKGROUND: Although airway eosinophilia is seen as a cardinal feature of asthma, data eosinophilia are still lacking on the proportion of the asthma group exhibiting raised airway eosinophilia. This study aimed to assess the distribution of sputum eosinophil count and its relationship with methacholine bronchial hyperresponsiveness in mild to moderate steroid-naive asthmatic people. METHODS: Sputum was induced by inhalation of hypertonic saline (NaCl 4.5%) in 118 mild to moderate steroid-naive asthmatic people consecutively recruited from our outpatient clinic, and in 44 healthy people. The asthma group was selected on the basis of an forced expiratory volume in 1 s (FEV(1)) of > or = 70% predicted, and a provocative methacholine concentration causing a fall of 20% in FEV(1) (PC20 methacholine; PC(20)M) < or = 16 mg/ml. RESULTS: In the asthma group, the median (range) of the percentage and the absolute values of sputum eosinophils were 4.8% (0-75) and 38 10(3)/g (0-14,191), respectively, vs 0% (0-2.3) (P < 0.001) and 0 10(3)/g (0-53) (P < 0.001) in healthy participants. Based on the 95% percentile for normal values calculated from our healthy group, 69% of the asthma group had significantly raised sputum eosinophil count (that is > 2%). In the asthma group, multiple regression analysis followed by a stepwise procedure revealed that sputum eosinophil count was significantly and inversely associated with PC(20)M accounting for 16% of its total variance (P < 0.001) while neutrophil counts positively related to PC(20)M accounting for 4% of total variance (P < 0.05). By contrast, no significant relationship was found between either eosinophil or neutrophil counts and the slope of forced vital capacity (FVC) vs FEV(1) from the methacholine challenge. CONCLUSIONS: We conclude that two-thirds of people in the mild to moderate asthma group had increased sputum eosinophilia, which plays a limited role in determining the degree of methacholine airway hyperresponsiveness.