Within-breath oscillometry for identifying exercise-induced bronchoconstriction in pediatric patients reporting symptoms with exercise

Background Evaluating oscillometry parameters separately for the inspiratory and expiratory breath phases and their within-breath differences can help to identify exercise-induced bronchoconstriction (EIB) in pediatric outpatients disclosing exercise-induced symptoms (EIS). Aims To assess the response in impedance parameters following an exercise challenge in patients reporting EIS. Methods Sixty-eight patients reporting EIS (34 asthmatics and 34 suspected of asthma, age mean = 10.8 years, range = 6.0–16.0) underwent an incremental treadmill exercise test. Spirometry was performed at baseline and 1, 5-, 10-, 15-, and 20-min post exercise. Oscillometry was performed at baseline and at 3- and 18-min post exercise. Bronchodilator response to 200 µg albuterol was then assessed. EIB was defined as a forced expiratory volume in 1 s (FEV1) fall ≥10% from baseline. Expiratory and inspiratory resistance (Rrs) and reactance (Xrs), their z-score (Ducharme et al. 2022), and their mean within-breath differences (ΔRrs = Rrsexp-Rrsinsp, ΔXrs = Xrsexp-Xrsinsp) were calculated. Receiver operating characteristic (ROC) curves and their areas (AUCs) were used to evaluate impedance parameters’ performances in classifying EIB. Results Asthmatic patients developed EIB more frequently than those suspected of asthma [18/34 (52.9%) vs. 2/34 (5.9%), p < 0.001]. In the 20 subjects with EIB, Rrsinsp, Rrsexp, Xrsinsp, and Xrsexp peaked early (3’), and remained steady except for Xrsinsp, which recovered faster afterward. ΔXrs widened 18 min following the exercise and reversed sharply after bronchodilation (BD) (−1.81 ± 1.60 vs. −0.52 ± 0.80 cmH2O × s/L, p < 0.001). Cutoffs for EIB leading to the highest AUCs were a rise of 0.41 in z-score Rrsinsp (Se: 90.0%, Sp: 66.7%), and a fall of −0.64 in z-score Xrsinsp (Se: 90.0%, Sp: 75.0%). Accepting as having “positive” postexercise oscillometry changes those subjects who had both z-scores beyond respective cutoffs, sensitivity for EIB was 90.0% (18/20) and specificity, 83.3% (40/48). Conclusion Oscillometry parameters and their within-breath differences changed markedly in pediatric patients presenting EIB and were restored after the bronchodilator. Strong agreement between z-scores of inspiratory oscillometry parameters and spirometry supports their clinical utility, though larger studies are required to validate these findings in a broader population.


Introduction
A common complaint in pediatric outpatients visiting the pulmonologist is the limitation of physical activity, which they often refer to as dyspnea, cough, chest pain, or other less-well-defined symptoms elicited by exercise.These patients embrace a heterogeneous group with some having a history of recurrent respiratory symptoms, while others refer to exercise limitation as their only complaint (1,2).Consequently, children are less prone to get involved in sports with further physical and psychosocial drawbacks (1).Testing for exercise-induced bronchoconstriction (EIB) is suitable, even when an asthma diagnosis seems inconsistent with the clinical history (3)(4)(5).Conversely, the presence of EIB in yet-diagnosed asthmatic children suggests poor disease control (2,3).
Classical exercise airway challenges entail repeated spirometry at baseline and in a timely manner after exercise, with skilled patients being able to perform these maneuvers.In children, oscillometry, which is also called the forced oscillation technique (FOT) needing quiet breaths, is feasible and reproducible (6,7).Throughout input wave or pressure pulses at the mouth opening, oscillometry measures of respiratory resistance (Rrs) and reactance (Xrs) disclose contributions to the total respiratory impedance (Zrs) from proximal and peripheral airways and tissues, and their response to several bronchial stimuli (6,8).Children tolerate well the midwave 8 Hz input frequency and several reference values for Rrs and Xrs at this frequency are now available (9)(10)(11).
A practical use of the oscillometry parameters is to compute them separately for the inspiratory and expiratory phases and assess their changes during the breath cycle.This approach is spreading as it may increase sensitivity and specificity to specific pathologies and conditions (6,12).In particular, the within-breath difference between the mean inspiratory Xrs (Xrs insp ) and the mean expiratory Xrs (Xrs exp ) increases with the occurrence of tidal expiratory airflow limitation (13).This approach has been useful to assess adult patients with chronic obstructed pulmonary disease (COPD) and children with cystic fibrosis (13)(14)(15).In this regard, no studies in pediatric patients have sought to evaluate the withinbreath changes in reactance during EIB.Also, previous studies did not consider changes in z-scores of impedance parameters, but absolute or relative changes were reported.The use of z-scores may better evaluate the effect of exercise.
We hypothesized that EIB [as assessed by the maximal forced expiratory volume in 1 s (FEV 1 ) fall], other than inducing changes in oscillometry parameters, also induces wide within-breath differences in Xrs that lessen as the bronchial patency restores.In keeping, we aimed to assess the response in impedance parameters following an exercise challenge in patients reporting exercise-induced symptoms (EIS), either without or with a previous diagnosis of asthma.For this purpose, we assessed Rrs, Xrs, at 8 Hz, and their within-breath differences at baseline and after the exercise challenge.

Subjects
We enrolled 73 consecutive outpatients attending the pulmonary section of our Pediatric Unit (Sant'Andrea Hospital, Rome, Italy) for EIS, to perform a bronchial challenge with exercise.After excluding five subjects not fulfilling the inclusion criteria (below), the remaining 68 patients yielded two groups of 34 subjects each as follows: (1) EIS without a previous diagnosis of asthma ("suspected asthma"), and (2) EIS and asthma ("asthma").The suspected asthma group comprised new-attendant children reporting exercise limitation as their main complaint, often associated with rhinitis, without a clinical history of asthma.None of these patients had hoarseness, stridor, chocking, or abnormal spirometry (flattening of the inspiratory portion of the flow volume) suspected of exerciseinduced laryngeal obstruction (EILO).The asthma group included patients whose diagnosis was based on both clinical history and pulmonary function (responsiveness to bronchodilators or to bronchial provocation tests) and had at least a partly controlled level of symptom control (16).For all subjects, inclusion criteria were a baseline FEV1 ≥80% of the predicted value, absence of respiratory infections in the previous 4 weeks, avoidance of inhaled corticosteroids, leukotriene antagonists, and antihistamines in the last 10 days, long-acting beta-agonists in the last 24 h, short-acting in the last 12 h.Parents signed a written informed consent; the Ethical Review Board of our Institution approved the study.

Study design
During a single session, patients underwent a medical visit, and parents completed a questionnaire on respiratory health including current respiratory symptoms and asthma medications.The child yielded information on EIS such as cough, dyspnea, and chest pain/tightness.Patients followed skin prick testing, oscillometry, and the fraction of exhaled nitric oxide (FE NO ) measurements in this order before baseline spirometry and exercise testing.

Skin prick tests
Sensitization to common inhaled and food allergens was tested by skin prick tests (SPTs) (ALK-Abellò, Milan, Italy) including house dust mites (Dpt and Dpf), molds (Alternaria alternata), dog and cat fur, Parietaria judaica, mixed grass pollens, mixed tree pollens, egg, and milk.Positive and negative controls were Histamine 0.1 mg/mL and glycerol solution, respectively.Positive SPTs were defined by a wheal size of at least 3 mm after subtraction of the negative control.

Oscillometry
Impedance measurements were obtained by oscillometry at a mono frequency of 8 Hz using the Resmon Pro Full device (ResTech, Milano, Italy).Children were seated, connected to the bacterial filter mouthpiece (DAR TM , Covidien IIc, Mansfield, USA), and breathed normally wearing a nose clip whereas the technician supported their cheeks.Three baseline replicates from at least 10 breaths each, free of artifacts, with a coefficient of variance (CV) ≤15% for Rrs were accepted (6).Tidal volume (Vt), respiratory rate (RR), and minute ventilation (V'E) for each set of measurements were also recorded.Expiratory and inspiratory Rrs and Xrs were expressed both as absolute means and as z-scores using reference values derived from the same model of device that we used (11) according to recent recommendations for technical standards in oscillometry measurements (6).Impedance modulus (Zrs) was also computed for both inspiratory and expiratory phases.Mean withinbreath differences (Rrs exp -Rrs insp , Xrs exp -Xrs insp ) were calculated and reported as ΔRrs and ΔXrs, respectively.

FE NO
FE NO was measured with an electrochemical device (HyPair FENO, Medisoft Group, Sorinnes, Belgium).Children did singlebreath maneuvers at constant expiratory pressure and flow of 50 mL/s as recommended.Mean values from two measurements agreeing within 10% from at least three expiratory maneuvers were calculated (17).

Spirometry
Spirometry was performed with a Quark PFT (Cosmed Srl, Rome, Italy) according to the ATS/ERS guidelines (18).The best FEV 1 from three measurements was recorded.Dynamic volumes and flows were both percentages and z-scores of the predicted values (19).

Exercise testing
After baseline oscillometry and spirometry, patients underwent an incremental treadmill exercise test (20).Running was set at 6 km/h with a pendant of 10% until children reached the target heart rate of over 85% of the predicted maximum, calculated as 220-age years (21).Room temperature (20-24°C) and relative humidity (50%-60%) were kept stable during the procedure.After the exercise, subjects repeated spirometry at 1, 5, 10, 15, and 20 min, and oscillometry at 3 and 18 min.The lowest FEV 1 was taken to calculate its decline from the baseline; EIB was defined as the fall in FEV 1 ≥10% from baseline after exercise (4).
Following the 20-min spirometry, subjects received albuterol 200 µg (MDI with spacer); after a 15 min interval, they repeated oscillometry and spirometry, in this order.
Postexercise changes in impedance from baseline were calculated and reported depending on the parameter: the maximal rise in Rrs or the maximal decrease in Xrs from their respective baseline values.The changes were both percentages from baseline and differences from baseline z-scores.To calculate differences, z-scores for impedance parameters at every postexercise step were computed; then the z-scores for postexercise values were subtracted from their respective baseline values.

Statistical analysis
Continuous variables were assessed for normal distribution (K-S test); pulmonary function variables were expressed as both means ± SD and z-scores using reference values for spirometry (19) and impedance (11); categorical variables were given as numbers and percentages.Postexercise changes and bronchodilator responses were expressed as medians and interquartile ranges (IQRs).The Mann-Whitney U test was used for unpaired comparisons and the non-parametric ANOVA Friedman with post hoc Dunnett's test to assess differences between postexercise and baseline measurements.The Wilcoxon test was used for paired comparisons between postbronchodilator and last postexercise measurements.
Contingency tables (χ 2 with Fisher's correction) were applied to compare categorical variables.Receiver operating characteristic (ROC) curves were constructed to analyze the relationship between sensitivity (Se) and 1-specificity (Sp) for impedance measurements in classifying EIB.The areas (AUC) under ROC curves were computed and used to identify the oscillometry parameters that better classified EIB.The cutoff values were determined by maximizing the Youden index.We also computed the cutoff values that maximize the Youden index among the ones providing Sp >Se.Spearman's Rho tests were used to assess correlations between continuous variables.The SPSS software (Version 27; SPSS Inc., Chicago, IL, USA) was used for statistical analyses.Two-tailed p values of <0.05 were regarded as statistically significant.

Identifying EIB by oscillometry
In the pooled population, maximal percentage impedance changes correlated slightly better with maximal percentage changes in FEV 1 than those in FEF 25-75 after exercise, while the opposite was true after bronchodilation.All correlations improved by using maximal changes in Rrs and Xrs z-scores (Table 3).

Postexercise outcome differences according to oscillometry and spirometry
The relationship between EIB defined by spirometry and changes in z-scores of inspiratory oscillatory parameters above the cutoffs maximizing the Youden index is described in Figure 4.
Approximately 90% (18 subjects including 12 males) of the 20 spirometry-defined EIB subjects responded to exercise according to both z-scored Rrs insp and Xrs insp , whereas 10% (two boys) did not respond above these z-score cutoffs.Conversely, among the 48 subjects without EIB, eight (16.7%, all males) subjects responded to exercise according to both z-score Rrs insp and z-score Xrs insp , and 12 subjects (25.0%, nine males) responded in accordance to either z-score Rrs insp or z-score Xrs insp .These two subgroups had similar baseline lung function and body mass index (BMI) percentile to the remaining 28 (58.3%)non-responder subjects according to oscillometry and without EIB.

Discussion
Our findings support the utility of impedance measurements following an exercise challenge in outpatients reporting EIS.The occurrence of EIB raised Rrs, decreased Xrs, and widened the within-breath difference in Xrs; such a difference reversed close to that baseline after bronchodilation with albuterol.To our knowledge, evaluating oscillometry parameters in terms of zscores computed separately for inspiration and expiration and the within-breath analysis of reactance has not previously been described to assess the exercise challenge.Also, we added to the scarce data on exercise testing at 8 Hz (22), a suitable forcing frequency for the pediatric population.
Notwithstanding that exercise increases minute ventilation, preceding studies assessing breathing parameters in adults (23) and adolescents (24) have found uninfluential V'E (23) and changes in Vt, RR, and V'E on oscillometry measurements for assessing EIB.Neither did we find a relationship between changes in ventilation and those in oscillometry early after exercise; still, the rise in Vt and V'E did correlate slightly with the fall from baseline in Xrs exp (but not with Xrs insp ) 18 min after the exercise ended.

Oscillometry by EIB outcomes
Several studies have reported changes in respiratory impedance following exercise in subjects with asthma or asthma-like symptoms including young adults (23, 25, 26), children, adolescents (22,24,(27)(28)(29)(30)(31), and preschool children (32, 33).These studies showed larger dynamic impedance changes from the few minutes following exercise in subsets of asthmatic patients or   In our population of subjects reporting EIS, changes in oscillometry were not statistically significant in subjects without EIB.This is not unexpected, as there is a consistent percentage of subjects reporting EIS whose response to exercise cannot be objective by lung function tests (24,34).BD improved oscillometry parameters in this group; however, improvements remained within the cutoff for BD response (relative percent change and z-score change) of healthy children aged 2-13 years (10).
Time changes in oscillometry parameters in EIB (spirometry assessed) subjects reported by previous studies (23,24,26,27) showed an early worsening (at 5 min) of both Rrs and Xrs after exercise and a faster Xrs recovery than of Rrs.This suggests an earlier functional recovery of less proximal airways than of the relatively proximal airways.Our data hold up the early change in impedance parameters as shortly as 3 min after exercise and the slower Rrs recovery than Xrs in EIB subjects.
Both Rrs insp and Rrs exp remained high after 18 min post exercise.Xrs insp gradually recovered while Xrs exp remained persistently decreased, i.e., the Xrs recovery was supported from Xrs insp only.This implies that the within-breath difference (ΔXrs) increased from baseline to the late postexercise step.Xrs reflects communicating lung volume and compliance (35) and is affected by ventilation heterogeneity.It changes minimally during tidal breathing in healthy subjects while it may drop sharply during expiration in subjects with tidal peripheral airway collapse (35,36).Our data suggest that ΔXrs widens with acute airway obstruction, mostly peripheral, following exercise.These may be the results of different mechanisms.The development of flow limitation in some airways (13), the lower elastic recoil, and higher collapsibility of the small airways at lower volumes (37) leads to higher ΔXrs both directly and indirectly by increasing heterogeneity during expiration.Our data showed that increased ΔXrs was not the result of a further Xrs exp decrease but of a faster Xrs insp recovery from the early postchallenge step fall.Measuring within-breath Xrs components may provide additional information for the analysis of changes in lung mechanics after exercise.

Identifying EIB by oscillometry
Changes in spirometry and oscillometry parameters were correlated but not totally overlapped.
Changes in z-scores rather than percentages from the baseline better assessed postexercise and postbronchodilator changes as detected by spirometry.This was especially verified for Xrs as drawbacks from using percentages of Xrs are especially evident when values close to zero are recorded (6).
Changes in z-scores of inspiratory rather than in expiratory oscillometry parameters best predicted EIB.Therefore, we agree with other studies reporting that inspiratory impedance parameters indicate EIB better than expiratory (24,29).Conversely, in school-age children born extremely preterm, inspiratory impedance did not mirror EIB, suggesting a distinct prematurity-driven airway reactivity (38).
Percentage changes from baseline of Zrs insp performed best against spirometry (Tables 3, 4), supporting the idea that this parameter, including contributions of both Rrs insp and Xrs insp changes, can better describe the changes in lung condition (24).However, reference equations are not available for Zrs insp (and their intrabreath components) yet.Therefore, we cannot verify if using z-score Zrs insp can further improve agreement with spirometry.The maximum postexercise widening of ΔXrs was also related to EIB even if it provided lower AUC than postexercise changes in z-scores of Rrs insp or Xrs insp .Reference values of ΔXrs are not available.Reporting ΔXrs as a z-score may improve the correlation with spirometry parameters when assessing responses to the exercise challenge.
Previous cutoffs for oscillometry with spirometry-based bronchial challenges in children refer to the methacholine testing, which (using different forced frequencies and devices) varies from a 45%-70% increase in Rrs and a 50%-80% drop in Xrs (6).Instead, cutoffs for oscillometry with exercise testing are less well established.A recent study described a 27% increase in R5 and a 47% decrease in X5 as predictors of EIB, based on a fall in FEV 1 ≥10% in asthmatic adults (23).Another study disclosed an increase of 0.035 kPa/L/s in R5 after 5 min of exercise (and a 0.055 kPa/L/s as maximum increase) to distinguish asthmatic patients with vs. those without airway hyperresponsiveness to methacholine (26).The most recent study, in asthmatic children, reported a 14.1% increase in R5 within 30 min post exercise as a diagnostic of EIB (31).Differently, among adolescents with or without reported exercise-induced dyspnea, cutoffs were drawn from 95% oscillometry changes in those regarded as healthy.Such an approach identified z-score increases in Rrs of 0.68 and Xrs of 1.76 as cutoffs for EIB (24).Our cutoff providing Sp>Se for z-score Rrs insp is close to the one determined on healthy children (0.67 vs. 0.68), though our z-score Xrs insp at 8 Hz is lower than their change in Xrs5 (1.14 vs. 1.76), as expected using higher stimulation frequencies and considering only the inspiratory phase.

Postexercise outcome differences according to oscillometry and spirometry
There is increasing evidence that spirometry and oscillometry are not always in accordance as they describe lung mechanics under different conditions (6,12,37).Spirometry requires a high lung volume maneuver and FEV 1 reflects mainly airflow resistance of proximal airways (39).Oscillometry requires tidal breaths at resting lung volume, is also sensitive to peripheral airways, and is deemed complemental to spirometry for assessing airway responsiveness (6,12).Changes in inspiratory Rrs well document the attenuation effect of EIB in spirometry with deep inspirations in asthmatic children in contrast with healthy controls (29).Small degrees of airway narrowing in healthy children have been attributed to transient hyperemia of the airway wall leading to local edema (29), a mechanism that could explain why the impedance increased after exercise in some of our subjects without EIB.
The best agreement between postexercise changes in spirometry (EIB) and oscillometry was seen in subjects showing changes beyond their identified cutoffs in both z-scores of Rrs insp and Xrs insp .This supports the idea of the importance of considering changes in both Rrs and Xrs and correcting for anthropometric factors.However, as expected the two techniques did not lead to identical results because of different physiological lung conditions during the measurements and different difficulties in performing the measurements (6)(7)(8)12).Two subjects who developed EIB (both males with asthma) did not show large changes in either z-scores of Rrs insp or Xrs insp .On the contrary, large changes in these two parameters occurred in eight male subjects without EIB who had evenly distributed asthma diagnoses (4:4).Incongruent growth of the airway's caliber relative to the growth of lung parenchyma, known as airway dysanapsis (characterized by low FEV 1 /FVC despite normal FEV 1 and FVC) is relevant in children (39, 40).Boys have smaller airways than girls proportional to their lung growth (40,41); obesity is associated with airway dysanapsis in children with or without asthma, and this airway-lung incongruence is linked to worse asthma control (40).Our eight male subjects without EIB who had large changes in both z-scores of Rrs insp and Xrs insp yielded no different baseline lung function (including FEV 1 /FVC) nor BMI percentile than the remaining subjects without EIB.Because these eight subjects had more changes in inspiratory reactance than the other subjects without EIB, they could have peripheral airway obstruction unrevealed by spirometry.Gender differences in small airway physiology have been reported among asthmatic subjects, as larger methacholine-induced gas trapping was found in males than females (42).Our eight subjects also showed greater changes in spirometry than the non-responding subjects after BD.An alternative hypothesis is that these eight boys, having a high airway hyperresponsiveness, could be prone to smooth-muscle relaxation after deep inhalations from FVC maneuvers.If so, the effect of exercise is more evident during quiet breathing and easily identified by oscillometry as the maneuver required for spirometry partially reverses the changes in lung mechanics.

Limitations
We acknowledge the following limitations.We involved outpatients attending the pulmonology service for EIS in a reallife approach, thus, we are cautious about comparing results with other settings that also involve healthy controls.Still, oscillometry parameters drive their use for screening patients with EIS in similar outpatient settings.The mono frequency of 8 Hz that we used allows the exploration of both proximal and less proximal airways; thus, the changes in Rrs insp and Xrs insp evinced the involvement of different airway segments during EIB in our study.
Physiological ventilatory responses to exercise could influence our oscillometry measurements.However, the expiratory reactance was the sole parameter slightly influenced by Vt whereas no inspiratory impedance components, i.e., those better identifying EIB, correlated with the ventilatory parameters.
The broncho-motor effect of deep inspiration with forced expiratory maneuvers could also modify our measures (29,43); to minimize this effect we did oscillometry maneuvers before those of spirometry.However, repeated maximal inspiratory maneuvers after the exercise challenge might result in bronchoprotection leading to underestimating EIB, especially in asthmatic subjects whose response to deep inspirations depends on the degree of airway obstruction (43).
Our subjects ran without a facemask, despite it being preferred to inhale medical dry air for the exercise testing protocol (3,4); this might decrease the number of positive test outcomes.Because more than half of our asthmatic patients had EIB, the overall blunt effect of deep inhalations on the broncho-motor tone or the environmental conditions during the test was probably modest.
We calculated ΔRrs and ΔXrs as mean within-breath differences (13)(14)(15), rather than computing the difference between the end-expiratory and the end-inspiratory impedance (i.e., both at zero flow).Although this recently suggested approach (44) has been indicated as potentially more sensitive to changes due to its reduced susceptibility to changes in breathing flow, our data exhibited very limited dependence on breathing patterns, which further strengthens the credibility of our findings.
In conclusion, we found marked changes in oscillometry parameters and raised within-breath differences in reactance among pediatric outpatients with self-reported EIS presenting EIB.Baseline oscillometry values were restored after the bronchodilator.The high agreement of z-scores of inspiratory parameters with spirometry supports their use for clinical evaluation although larger studies are needed to confirm the validity of the identified cutoff in a more general population.More studies are needed to establish the clinical role of withinbreath differences in impedance during EIB in these children.

FIGURE 1 Z
FIGURE 1 Z-scores of impedance parameters measured at baseline, after the exercise challenge, and after bronchodilation (BD) in patients with exerciseinduced bronchoconstriction (EIB) and those without.(A) z-score Rrs insp , (B) z-score Rrs exp , (C) z-score Xrs insp , and (D) z-score Xrs exp .*p < 0.01 and **p < 0.001 for postexercise measurements (3 and 18 min) vs. baseline, or BD vs. 18 min after exercise.

FIGURE 4
FIGURE 4 Venn diagram on the 40 patients with abnormal postexercise changes as defined either by spirometry or by oscillometry: 20 patients with EIB (fall in FEV 1 ≥10%) and 20 patients without EIB and impedance changes beyond their cutoffs (n = 8: rise in Rrs insp z-score>0.41;n = 4: fall in z-score Xrs insp >−0.64; n = 8: both changes in Rrs insp and Xrs insp z-scores are abnormal).The other 28 patients (lower right corner), did not change by spirometry nor by oscillometry.See also maximized Youden index cutoffs for EIB of oscillometry parameters in Table4.

TABLE 1
Characteristics of the patients reporting exercise-induced symptoms (EIS).

TABLE 2
Maximal postexercise changes and bronchodilator responses (BDR) by EIB outcomes.
EIB: exercise-induced bronchoconstriction.Rrs, Xrs, and Zrs: respiratory resistance, reactance, and impedance modulus, respectively, measured during inspiration (insp) and expiration (exp).Maximal postexercise fall (or rise) changes from baseline are shown as both percentages and differences between z-scores at the respective steps.Bronchodilator responses were calculated from the last postexercise measurement.Variables are expressed as medians and interquartile ranges (IQRs).*p< 0.01.**p< 0.001 vs. subjects without EIB.§p < 0.05.

TABLE 3
Spearman's correlations between maximal changes in impedance and spirometry parameters after exercise testing and following bronchodilator administration in the pooled population (n = 68).
Rise (or fall) in z-scores was calculated as the difference between the maximal postexercise Rrs z-score (or the minimal postexercise Xrs z-score) and the respective baseline z-score.Bronchodilation was calculated as the difference between the postbronchodilator z-score and the postexercise (18 min) z-score for each oscillometry parameter.Postexercise and bronchodilator responses are also shown as percentage changes in respiratory resistance (Rrs), reactance (Xrs), and impedance modulus (Zrs), both during inspiration (insp) and expiration (exp).

TABLE 4 EIB
(fall in FEV 1 ≥10%) by cutoff values of impedance for all the patients in the group (n = 68).
Rise (or fall) in z-scores: Difference between the maximal postexercise Rrs z-score (or the minimal postexercise Xrs z-score) and their respective baseline z-score.ΔXrs: Difference between mean expiratory and inspiratory reactance (cmH 2 O × s/L) measured 18 min after the exercise challenge.PPV, positive predictive values; NPV, negative predictive values.*PPV and NPV are computed considering the EIB incidence of our study (29%).