Cut‐off values to evaluate exercise‐induced asthma in eucapnic voluntary hyperventilation test for children

Abstract Background and Aim The eucapnic voluntary hyperventilation (EVH) testing is a diagnostic tool for diagnostics of exercise‐induced bronchoconstriction; while the testing has become more common among children, data on the test's feasibility among children remain limited. Our aim was to investigate EVH testing feasibility among children, diagnostic testing cut‐off values, and which factors affect testing outcomes. Methods We recruited 134 patients aged 10–16 years with a history of exercise‐induced dyspnoea and 100 healthy control children to undergo 6‐min EVH testing. Testing feasibility was assessed by the children's ability to achieve ≥70% of the target minute ventilation of 30 times forced expiratory volume in 1 s (FEV1). Bronchoconstriction was assessed as a minimum of 8%, 10%, 12%, 15% or 20% fall in FEV1. Patient characteristics were correlated with EVH outcomes. Results Overall, 98% of the children reached ≥70%, 88% reached ≥80%, 79% reached ≥90% and 62% reached ≥100% of target ventilation in EVH testing; of children with a history of exercise‐induced dyspnoea, the decline percentages were as follows: 24% (≥8% fall), 17% (≥10% fall), 10% (≥12% fall), 6% (≥15% fall) and 5% (≥20% fall) in FEV1, compared to 11%, 4%, 3%, 1% and 0% among the healthy controls, respectively. Healthy controls and boys performed testing at higher ventilation rates (p < .05). Conclusion Eucapnic voluntary hyperventilation testing is feasible among children aged 10–16 years and has diagnostic value in evaluating exercise‐induced dyspnoea among children. A minimum 10% fall in FEV1 is a good diagnostic cut‐off value. Disease status appears to be important covariates.

The American Thoracic Society (ATS) has recommended that EIB should be diagnosed by establishing changes in lung function provoked by exercise (Parsons et al., 2013). The eucapnic voluntary hyperventilation (EVH) test is an alternative method to other indirect or direct bronchial challenge tests such as exercise challenge or methacholine challenge test that has been described as a sensitive technique for diagnosing EIB (Anderson, Argyros, Magnussen, & Holzer, 2001;Dickinson, McConell, & Whyte, 2011). The EVH test has traditionally been used for elite athletes (Anderson et al., 2001;Dickinson et al., 2011) and is widely regarded as the gold standard tool for assessing EIB among athletes (Hull, Ansley, Price, Dickinson, & Bonini, 2016). A minimum 10% fall in forced expiratory volume in 1 s (FEV1) is generally considered significant (Hallstrand et al., 2018;Parsons et al., 2013). There is only one large scale study in ordinary adults (Brummel, Mastronarde, Rittinger, Philips, & Parsons, 2009), where 71% of adults reached minimum 70% of target minute ventilation, meaning 60% of maximal minute ventilation. In total, 28% of the study patients with asthma-like symptoms had 10% fall of FEV1%. On the other hand, 44 of 224 (20%) non-symptomatic adult elite athletes had minimum 10% fall of FEV1 after EVH . According ERS specificity is higher with criterion of minimum fall of 15% FEV1 compared cut-off 10% (Hallstrand et al., 2018). EVH can also provoke vocal cord dysfunction (Christensen & Rasmussen, 2013;Turmel, Gagnon, Bernier, & Boulet, 2015 EVH testing is becoming more common among children for the diagnostics of exercise-induced dyspnoea, but the data remains scarce. In this study, we thus aimed to explore the feasibility of EVH testing among children with exercise-induced dyspnoea. We hypothesized that EVH testing would be feasible among children aged 10-16 years and that the test could provoke bronchoconstriction among children who experience exercise-induced dyspnoea. We also wanted to determine whether a cut-off value of a 10% fall in FEV1 could be used among children, much as such a cut-off value is recommended for adults. Finally, we have investigated whether patient characteristics might influence testing outcomes.

| Recruitment
The study was conducted at the paediatrics departments of the university hospitals of Turku and Kuopio, Finland. The inclusion criteria included a suspicion of pathological reasons for exercise-induced dyspnoea, exercise-induced bronchoconstriction or dysfunctional breathing in patients between 10 and 16 years. The exclusion criteria were physical inactivity, severe comorbidity or chronic autoimmune disease. The 100 healthy controls, from the same age range, were recruited through local sports clubs. Their inclusion criteria included engaging in sporting activity without symptoms of exercise-induced dyspnoea, active asthma or severe comorbidity. The Ethics Committee of the Hospital District of Southwest Finland approved the study, and written informed consent was provided by all the participants and their guardians.

| Background data
The Childhood Asthma Control Test was completed by the participants and their guardians (Liu et al., 2007). In addition, a written questionnaire, completed by the guardian, was used to collect information about the subjects' previous medical history including doctor diagnosed asthma ever, current sporting activity, allergies, and acute and chronic respiratory symptoms, including cough, exerciseinduced dyspnoea, running nose, fever and throat symptoms.

| Testing prerequisites
Beta 2 -agonists were not administered for 12 hr before the tests. The baseline FEV1 had to be at least 70% of the age-and height-related reference values (Koillinen, Wanne, Niemi, & Laakkonen, 1998). If a patient had an acute respiratory infection, then the test was postponed for 2 weeks.

| Flow-volume spirometry
The test began with baseline spirometry in which FEV1 was the main outcome (Moore, 2012). The subjects then underwent EVH testing, as described below, and spirometry was repeated 1, 5 and 10 min after the test. Finally, patients were given 0.4 mg of salbutamol in the form of a dry powder at the Turku Centre (Buventol Easyhaler: Orion Pharma) and as a spray at the Kuopio Centre (Ventoline Evohaler: Glaxo Wellcome Production) with a Babyhaler spacer device (Glaxo Wellcome Production), based on each centre's routine clinical practice. The spirometry test was repeated 15 min after the administration of salbutamol. During bronchodilatation testing, an improvement of 12% or more in FEV1 compared to baseline was interpreted as significant (Pellegrino et al., 2005).

| Eucapnic voluntary hyperventilation test
The duration of the EVH test was 6 min (Anderson et al., 2001;Burman et al., 2018). The test equipment is shown in Figure 1. The test involves the patient inhaling gas that contains oxygen plus 74% nitrogen and 5.1% carbon dioxide. The target minute ventilation in the EVH test was defined as 30 times each patient's baseline FEV1, which was equivalent to 85% of minute ventilation volume (Hallstrand et al., 2018;Parsons et al., 2013). The mouthpiece used Medical Inc. The feasibility of the EVH test was assessed by the ability of participating children to achieve the minimum 70% target level of the minute ventilation volume (Hallstrand et al., 2018;Parsons et al., 2013). Minute ventilation was measured using the mouthpiece airflow sensor in real time (Burman et al., 2018). The software (WinCPRS, Absolute Aliens Oy) used in the EVH test equipment allowed us to show in real time the continuous 10-s sliding average of the minute ventilation on the monitor graphically with the specific target level-line that allowed the study subject to maintain the targeted minute ventilation. At the end of the test, all ventilation data were saved, and the average minute ventilation was calculated for the 6-min examination time.

| Outcomes
The primary aim of this study was to investigate the feasibility of EVH testing by assessing whether participants could achieve ≥70% of the target minute ventilation; additional target levels of ≥80%, ≥90% and ≥100% were also analysed. The second aim was to evaluate whether a guideline-based cut-off value of a 10% fall in FEV1 (Hallstrand et al., 2018;Parsons et al., 2013) provoked by hyperventilation could differentiate cases from controls; additional target levels of 8%, 12%, 15% and 20% fall in FEV1 were also analysed.
The third aim was to determine whether common patient characteristics, age, sex, current physician-diagnosed asthma, Childhood Asthma Test score (Liu et al., 2007), current atopic eczema, baseline FEV1, achieved minute ventilation level (70%-99% vs. ≥100% level) or response to bronchodilator correlated with the EVH outcomes. were used. Logistic regression analysis was used to adjust for age and sex during analyses of target minute ventilation between groups.

| Statistical analysis
The statistical significance was established at p < .05.

| Study population
We enrolled 234 children; of these, 134 were cases with a history of exercise-induced dyspnoea, and 100 were controls without any exercise-induced symptoms.

| Subject characteristics
The mean age of the 234 children was 13.7 years (standard deviation [SD] 1.9 years), and the mean baseline FEV1 was 96% of the predicted normal level (Table 1). Girls achieved a greater per cent predicted FEV1 than boys: (98.7% [SD 11.0] vs. 94.0% [SD 11.6]; p = .002). Cases with a history of exercise-induced dyspnoea were more often girls than among the controls (57% vs. 29%; p < .001), had more atopic eczema (33% vs. 19%; p = .018), more often had physician-diagnosed asthma (33% vs. 5%; p < .001) and had lower Childhood Asthma Test scores than the healthy controls (mean 21.1 vs. 26.2 points; p < .001) ( Table 1). The controls participated in competitive sporting activities more than the patients (p < .001), possibly due to fact that they were recruited from sports groups.
No other differences in characteristics were found between the subjects (Table 1).

| Ability to maintain target minute ventilation during eucapnic voluntary hyperventilation testing
Of all 234 children, the minimum 70% of target minute ventilation was achieved by 229 (98%), ≥80% by 207 (88%), ≥90% by 185 (79%) and ≥100% by 144 (62%) of subjects. The boys achieved a mini- ; univariable p = .001; sex-adjusted p = .017), which were generally more positive than the cases, but after adjusting for sex, the significance in which a minimum of 90% of the target was reached was lost ( Figure 2). Interestingly, all 26 children who experienced bronchoconstriction reached a minimum 70% of the target minute ventilation.

| Fall in forced expiratory volume in 1 s testing after eucapnic voluntary hyperventilation testing
Overall, the mean fall in FEV1 among all children was −4.9% (SD 5.7%). Bronchoconstriction, assessed as a minimum 8% fall in FEV1, occurred among 44 (19%) of all 234 children; other rates included a minimum 10% fall in FEV1 26 (11%), a minimum 12% fall in FEV1 16 (6.8%), a minimum 15% fall in FEV1 9 (3.8%) and a minimum 20% fall in FEV1 7 (3.0%). Age, sex and diagnosis of atopic eczema or asthma did not affect the fall in FEV1 after the EVH (data not shown).
A greater fall in FEV1 was observed among the cases than among the controls (mean −6.0% vs. −3.6%; p = .010); see Figure 3. Cases had more bronchoconstriction than controls among every cut-off  Figure 3).
No differences were noted among children who reached 70%-99% of the target minute ventilation or children who reached 100% of the target in the bronchoconstriction-related findings (Table 2).
There was no correlation between reaching target minute ventilation volume and fall of FEV1 after EVH, (Spearman correlation, r = .061; p = .359).

| Cut-off values in clinical decision-making
When we used a cut-off value of an 8% fall in FEV1, 32 of 43 (74%) children were cases, while using a cut-off value of 10% fall in FEV1, 22 of 26 (85%) children were considered cases. The proportions in cut-off values of 12% (81%) or 15% (89%) of cases were similar compared to the cut-off value of 10%.

TA B L E 1 Baseline characteristics of the children
If a cut-off value of 12% had been used instead of a 10% value, then there would be a 41% reduction in positive findings among the cases. A cut-off value of 15% had a 64% reduction in positive findings among the cases, and a 20% cut-off had a 68% reduction in positive findings among the cases (Figure 3).

| Sensitivity and specificity in different cut-off values to identify exercise-induced asthma
Sensitivity and specificity in different cut-off values are shown in Table 3.

| Dysfunctional breathing
Of the 134 cases, 16 (12%) had objective symptoms, such as inspiratory stridor, hyperventilation or other breathing abnormalities, without significant fall of FEV1 and a lack of bronchodilator response.
None of the healthy controls experienced dysfunctional breathing during EVH testing (p < .001).

| D ISCUSS I ON
Three main results arose from this study. First, most of the 10to 16-year-old children successfully conducted the EVH testing F I G U R E 2 Reaching target minute ventilation volume in EVH testing. The black column presents the proportion of cases; the grey column presents the proportion of controls; *p < .05. Target minute ventilation volume defined as 30 times forced expiratory volume in 1 s F I G U R E 3 Proportion (%) of children having bronchoconstriction in different cut-off values after EVH testing. The black column presents the proportion of cases; the grey column presents the proportion of controls; *p < .05; **p < .01. FEV1: forced expiratory volume in 1 s without any side effects, and 70% of the target may be considered an acceptable ventilation rate. The real-time aid of graphical and visual feedback was useful in maintaining ventilation rates. Second, a cut-off 10% fall in FEV1 is useful for identifying those patients with exercise-induced bronchoconstriction, and bronchodilatation testing rarely appeared to be positive after EVH testing when using a cut-off of 12% from the baseline. Third, EVH testing is useful in identifying cases with dysfunctional breathing.
Almost all children were able to complete the EVH test at the 70% target level, regardless of any symptoms they may have experienced during the test. The subjects tolerated the EVH test well, and no additional side effects except some couching due to increased mucus production were observed among the participants.
Interestingly, at high (90%-100% of the target) ventilation rates healthy children had better performance than cases. In contrast, many previous studies have shown that only the minority (range 0%- ). In addition, among the general adult population, 70% of the target was achieved at a much lower rate (71%) than in our study (Brummel et al., 2009). The graphical real-time biofeedback signal data, which enabled the children to regulate their ventilation effortlessly, might have played a role in the positive results. Another key success factor might have been the research personnel's encouragement during the testing.
Children with previous asthma diagnosis had no more bronchoconstriction than children without asthma diagnosis in early childhood. This could be because many children with clinical diagnosis of asthma in early childhood had actually suffered from virus-induced wheezing rather than "real asthma." In our study, a 10% cut-off in the fall in FEV1 after EVH testing was considered optimal for the diagnosis of bronchoconstriction, because it most strikingly differentiated cases from controls. Cut-offs of ≥10 to ≥20% had specificity to identify exercise-induced asthma at 98%-99%, but they also markedly decreased the sensitivity from 51% to 16%, implying that if the cut-off minimum 20% instead of 10% is used, the sensitivity would decrease significantly. Ventilation rate had no influence on this difference, which further supports the target minute ventilation of ≥70% during EVH testing among 10-to 16-year-old children. Our results are in agreement with current recommendations, according to ATS and ERS (Hallstrand et al., 2018;Parsons et al., 2013). greater per cent predicted FEV1, which was probably the explanation for the difference. Age, diagnosis of atopic eczema or asthma, baseline FEV1 or fall in FEV1 after EVH did not affect the results we obtained in our study. In a previous study, a fall in FEV1 did not affect the reaching of target minute ventilation (Chateaubriand do Nascimento Silva Filho et al., 2015). Previous studies' potential confounding factors have not usually been reported.
To our knowledge, this was the largest study using EVH testing to have been conducted with children, which is a major strength of the study. One limitation of the study is that spirometry follow-up after EVH was not made according ERS recommendations every 3 min. First spirometry follow-up was made 1 min after EVH and fatigue might have affected first spirometry obtained. Another limitation of our study was that both cases and control children were actively engaged in sports. Their target minute ventilation achievements during EVH testing may not have been as good if the controls had been physically inactive. Our results are thus generalizable only to those who are active in sports. The proportion of males among the controls was higher compared to the cases, but sex-adjusted analyses showed that the findings were independent of sex.
We found that EVH testing is feasible for 10-to 16-year-old children. The reaching of a minimum 70% of target minute ventilation volume may be considered acceptable performance. Another finding was that a cut-off value of a minimum 10% fall in FEV1 also works well among children. EVH testing is also useful in identifying cases with dysfunctional breathing. Our data provide important evidence for the current ERS and ATS guidelines (Hallstrand et al., 2018;Parsons et al., 2013).