Results
Seventy-two asthmatics (44 boys) and 29 healthy children (13 boys), aged 8 to 11 years, were included in this study. 49(68%) asthmatic children had no symptoms during the last 3 months while 23 patients (32%) had symptoms as chronic dry cough in 14(61%) or cough/wheezing on exertion in 9(39%). Thirty-two patients (44%) had no anti-asthmatic treatment while 40(56%) were treated by inhaled steroids alone in 10 (14%), or associated with leukotriene antagonist in 2 (3%), with long-acting BD agent in 9 (13%), or both in 9 (13%). 10 (14%) asthmatics had leukotriene antagonist alone.
Patients’ characteristics and spirometric parameters at baseline, expressed as absolute value or z-score, and their BD responses expressed as Δ%, are presented in table 1 . The two populations were similar for age, height, weight and body mass index. FVC and zFVC were not different between the two groups. Asthmatics had significantly smaller absolute values and z-scores for FEV1 (p=0.003 , respectively p=0.04 ), FEV1/FVC (p=0.002 for both ), FEF25-75% (p=0.0003, respectively, p=0.004 ), FEF50/PEF (p=0.01 , respectively p=0.007 ) and β-angle (p=0.009 , respectively p=0.006 ) than healthy children. More asthmatics, 32 (44%), than controls, 7 (24%), had β-angle <180° or FEF50/PEF <0.05 that define a concave shape of FVL but without statistical significance. 10 (14%) asthmatic children had zFEV1/FVC <-1.64 associated with zFEV <-1.64 in 6 (8%) of them.
Spirometric parameters after BD inhalation, expressed in z-score, were lower in asthmatics than controls but without statistical significance, except for zFEV1/FVC-bd (p=0.02 ) (supplementary table).
BD responses showed no difference for Δ%FEV1 between asthmatic and control children. Asthmatics presented higher BD responses compared to controls statistically significant for Δ%FEV1/FVC (p=0.04 ) and Δ%FEF50/PEF (p=0.02 ), but not for Δ%β-angle (p=0.07).
The comparison between the 23 (32%) asthmatic children presenting respiratory symptoms during the last 3 months and the 49 (68%) asymptomatic patients is presented in table 2 . Age, weight and height were comparable in both groups. zFEV1 was similar and zFEV1/FVC, zFEF25-75%, zFEF50/PEF and zβ-angle were lower in the group of children with symptoms vs no-symptoms but without statistical significance. BD change was not different between the 2 groups for Δ%FEV1 and Δ%FEV1/FVC but it was significantly higher in asthmatics with symptoms for Δ%FEF50/PEF and Δ%β-angle, reported to the asymptomatic group (p=0.02 ).
ROC curves are represented for the different parameters at baseline in z-score in figure 2 and for their BD reversibility infigure 3 . AUC (95CI), sensitivity, specificity and respective cut-offs are presented in table 3 for the parameters at baseline in z-score, for their BD reversibility expressed as Δ%, and their combination.
At baseline, zFEV1/FVC, zβ-angle or zFEF50/PEF presented similar AUC’s, slightly more elevated than zFEV1’s AUC (figure 2 ). The spirometric parameters exhibited different patterns of sensitivity and specificity, i.e. zFEF50/PEF and zβ-angle showed the best sensitivity while specificity was similar for zFEV1, zβ-angle or z-FEF50/PEF (table 3 ). When combining the spirometric parameters, the AUC’s improved and best sensitivity was obtained when considering zFEV1+zβ-angle or zFEV1+z-FEF50/PEF, at the respective thresholds, while best specificity was obtained for zFEV1/FVC+zFEF50/PEF (table 3 ).
ROC curves for BD reversibility showed larger AUC for Δ%FEF50/PEF, followed by Δ%FEV1/FVC and Δ%β-angle (figure 3 and table 3 ). Sensitivity was best for Δ%β-angle or Δ%FEF50/PEF and very low for Δ%FEV1 (table 3 ). Specificity was best for Δ%FEV1 and low for Δ%β-angle. The association of parameters improved the AUC’s but not the sensitivity – specificity.