Original Research

Unsupervised field-based exercise challenge tests to support the detection of exercise-induced lower airway dysfunction in athletes

Abstract

Background Athletes are at risk for developing exercise-induced lower airway narrowing. The diagnostic assessment of such lower airway dysfunction (LAD) requires an objective bronchial provocation test (BPT).

Objectives Our primary aim was to assess if unsupervised field-based exercise challenge tests (ECTs) could confirm LAD by using app-based spirometry. We also aimed to evaluate the diagnostic test performance of field-based and sport-specific ECTs, compared with established eucapnic voluntary hyperpnoea (EVH) and methacholine BPT.

Methods In athletes with LAD symptoms, sensitivity and specificity analyses were performed to compare outcomes of (1) standardised field-based 8 min ECT at 85% maximal heart rate with forced expiratory volume in 1 s (FEV1) measured prechallenge and 1 min, 3 min, 5 min, 10 min, 15 min and 30 min postchallenge, (2) unstandardised field-based sport-specific ECT with FEV1 measured prechallenge and within 10 min postchallenge, (3) EVH and (4) methacholine BPT.

Results Of 60 athletes (median age 17.5; range 16–28 years.; 40% females), 67% performed winter-sports, 43% reported asthma diagnosis. At least one positive BPT was observed in 68% (n=41/60), with rates of 51% (n=21/41) for standardised ECT, 49% (n=20/41) for unstandardised ECT, 32% (n=13/41) for EVH and methacholine BPT, while both standardised and unstandardised ECTs were simultaneously positive in only 20% (n=7/35). Standardised and unstandardised ECTs confirmed LAD with 54% sensitivity and 70% specificity, and 46% sensitivity and 68% specificity, respectively, using EVH as a reference, while EVH and methacholine BPT were both 33% sensitive and 85% specific, using standardised ECTs as reference.

Conclusion App-based spirometry for unsupervised field-based ECTs may support the diagnostic process in athletes with LAD symptoms.

Trial registration number NCT04275648.

What is already known on this topic?

  • Exercise-induced lower airway narrowing may result in lower airway dysfunction (LAD), a highly prevalent condition in endurance athletes, particularly in winter and pool-based sports.

  • The lack of association between clinical symptoms of LAD and detection of exercise-induced lower airway narrowing in athletes requires an objective bronchial provocation test (BPT) to confirm the diagnosis of LAD.

  • An exercise challenge test (ECT) would presumably be the best way to detect exercise-induced lower airway narrowing and confirm the LAD diagnosis. Still, ECTs are resource-demanding and challenging to standardise.

  • Laboratory tests, including the eucapnic voluntary hyperpnea (EVH) and the methacholine BPT, are widely used as surrogates to ECTs to document LAD in athletes.

What this study adds

  • Using app-based spirometry for unsupervised field-based ECTs may detect lower airway narrowing in more athletes than EVH and methacholine BPT.

  • In athletes with symptoms suggestive of LAD, unsupervised field-based ECTs using app-based spirometry may facilitate the diagnostic process.

How this study might affect research, practice and/or policy

  • App-based spirometry may aid the diagnostic process of athletes with LAD symptoms.

  • The app-based spirometry is easily available and has low cost, which may reduce the worldwide gap in access to appropriate assessment of athletes with LAD symptoms.

Introduction

Athletes are at risk for developing exercise-induced lower airway narrowing.1 2 This condition is mainly caused by a release of local inflammatory mediators (eg, prostaglandins, leukotrienes, histamine) in response to evaporative water-loss from the airway surface3 4 and may be exacerbated by dry or cold air,5 or inhaling allergens.3 Symptoms suggestive of exercise-induced lower airway narrowing are excessive mucus, wheezing, cough, chest tightness and dyspnoea.6 Several clinical definitions have been used interchangeably for this condition, for example, exercise-induced asthma, exercise-induced bronchoconstriction and airway hyper-responsiveness, prompting the need for the recently introduced collective term, lower airway dysfunction (LAD).2 7

A correct diagnosis of LAD is important to normalise lung function and enable exercise without symptoms.8 In athletes, symptoms of LAD are poorly associated with detecting lower airway narrowing by an objective bronchial provocation test (BPT) to confirm the diagnosis.9–11 Thus, the International Olympic Committee-Medical Commission (IOC)8 and the World Anti-Doping Agency12 refer to Global Initiative for Asthma13 and strongly suggest the presence of both LAD symptoms and a positive BPT before appropriate LAD therapy is initiated in this group. Presumably, the best way to detect exercise-induced lower airway narrowing would be to assess lung function before and after a relevant training session, that is, an exercise challenge test (ECT).14 However, ECTs are resource-demanding in healthcare personnel and equipment, and challenging to standardise in terms of ambient conditions, exercise duration and workload.15 16 Therefore, laboratory BPTs are often used as surrogate diagnostic tests for ECT. Direct BPTs stimulate bronchial smooth muscle contraction to provoke airway narrowing, most commonly by inhalation of methacholine.4 17 In indirect BPTs, airway narrowing is induced by mimicking the pathophysiology of exercise through eucapnic voluntary hyperpnoea (EVH), or by inhaling strategies using mannitol or nebulised adenosine 5′-monophosphate. Based on its high specificity in confirming LAD, EVH has been suggested as the ‘gold standard’ for diagnosing LAD in athletes.8 13 However, a recent systematic review from an IOC consensus group14 showed that if ECTs are field based and sport specific, they seemed to have similar specificity and somewhat higher sensitivity than EVH to confirm LAD.

In the present paper, the objective was to assess if the novel concept of including a wireless, hand-held turbine-spirometer connected to a cloud-based smartphone application for unsupervised field ECTs, could aid the diagnostic process in athletes with LAD symptoms. First, we aimed to assess if unsupervised field-based ECTs could confirm LAD using app-based spirometry. Second, we aimed to evaluate the diagnostic test performance of field-based ECTs compared with EVH and methacholine BPT.

Methods

Study design

In this open diagnostic study examining athletes with LAD symptoms, all participants went through, in random order, at least one standardised field-based ECT, at least one unstandardised field-based sport-specific ECT, an EVH and a methacholine BPT, on four different days within 4 weeks (figure 1). In the last 6 hours before BPTs, participants were instructed to avoid strenuous exercise, heavy meals, caffeine-containing food or beverage, and nicotine. Any use of beta2-agonists, ipratropiumbromide or other asthma medication was discontinued before all BPTs, in line with the European Respiratory Society (ERS) guidelines.18

Figure 1
Figure 1

Flow chart showing the bronchial provocation tests included in the study.

Study population

Athletes aged 16–50 years, with LAD symptoms (ie, excessive mucus, wheezing, cough, chest tightness, dyspnoea) for >3 months and with current symptoms the last 8 weeks, were recruited from elite sport high schools, the website of The Norwegian School of Sport Sciences and through coaches employed at the National Olympic Training Centre (Olympiatoppen) in Oslo, Norway. Only athletes who were able to comply with the study procedures were enrolled. Exclusion criteria were forced expiratory volume in 1 s (FEV1)<75% or a forced vital capacity <80% according to the Global Lung Function Initiative (GLI 2012),19 oral corticosteroids or infections in the last 3 months, chronic disease and pregnancy.

Data collection

Electronic questionnaire

Participating athletes received a structured electronic questionnaire by mail on demographic background factors, respiratory health and treatment history, and healthcare utilisation at study enrolment.

Lung function

Dynamic spirometry for the EVH or methacholine BPT was measured by maximal expiratory flow-volume curves (MasterScreen Pneumo Jäger, Würzburg, Germany)20 according to standardised guidelines from ERS and American Thoracic Society (ATS).18 21

For the ECT-related spirometries, the athletes downloaded the CE-marked AsthmaTuner app on their iOS or Android smartphone (MediTuner AB, Stockholm, Sweden). The cloud-based system of AsthmaTuner is described previously,22 and includes (1) app modules for standardised and unstandardised ECTs connected to a (2) hand-held validated Bluetooth turbine-spirometer (MIR Spirobank Smart, Rome, Italy).23 24 Spirometry results are presented to the clinician’s web interface in real time.25 The software performs quality controls of lung function manoeuvres in line with the built-in ATS/ERS guidelines, ensuring that the FEV1 is based on a maximal effort and requires a minimum of three acceptable manoeuvres for an approved test.21 Reference values for the Jäger and the MIR Spirobank were calculated according to GLI 2012.19 The athletes were educated on performing a lung function test on the spirometer at inclusion, performing several tests under supervision to ensure proper execution.

Standardised field-based exercise test

Participants performed at least one representative (ie, provoked LAD symptoms) standardised field-based ECT supervised by the app. They were instructed to repeat the ECT if no LAD symptoms occurred to reduce the risk of false negative results. Ambient conditions and age-predicted maximal heart rate (HR) were registered in the app prechallenge. To secure standardisation, no warm-up was allowed before high-intensity exercise requiring ≥85% of maximal HR for 8 min. HR was measured using an HR-monitor and registered in the app immediately postchallenge. Measurements of FEV1 were conducted prechallenge and 1 min, 3 min, 5 min, 10 min, 15 min and 30 min postchallenge using the app-based spirometer. The ECT was deemed positive by a ≥10% drop in FEV1 at two consecutive time points postchallenge.18 Perceived LAD symptoms were registered at the same time points postchallenge to capture information about symptom presence.

Unstandardised field-based sport-specific exercise test

Participants performed at least one representative unstandardised (ie, any sport or competition) field-based sport-specific ECT and were instructed to repeat the ECT if no LAD symptoms occurred. The app-based spirometer was used prechallenge and at least once within 10 min postchallenge. The ECT was deemed positive by one single ≥10% drop in FEV1 postchallenge, in line with most previous papers including such unstandardised field-based ECTs.26–32

Eucapnic voluntary hyperpnoea

According to standardised protocols, a minute ventilation (VE) at a target of 85% of maximal voluntary ventilation (FEV1X30) was secured by continuously measuring expired flow throughout 8 min EVH, the same duration as for standardised ECT.12 33 34 The athletes breathed a gas mixture of 21% O2, 5% CO2 and 71% N2 from two gas cylinders via a commercial system (EucapSys SMTEC, Switzerland). The FEV1 was recorded prechallenge and 1 min, 3 min, 5 min, 10 min, 15 min, 20 min, 25 min and 30 min postchallenge, and the EVH was deemed positive by a ≥10% drop in FEV1 at two consecutive time points postchallenge.18

Methacholine BPT

According to ATS guidelines,32 inspiration-triggered nebuliser (MasterScreen Pneumo Jäger, Würzburg, Germany) delivered methacholine in doubling doses until a ≥20% drop in FEV1 was observed. Linear interpolation on the semilogarithmic dose-response curve was used to calculate the accumulated provocation dose (PD20) causing 20% drop in FEV1, and the methacholine BPT was deemed positive by PD20<4 µmol.35

Definitions and outcomes

We defined a confirmed LAD diagnosis as a positive BPT, that is, standardised ECT, unstandardised ECT, EVH or methacholine test, in athletes with a history of LAD symptoms.

Athletes without a confirmed LAD diagnosis were referred to further clinical examinations if indicated based on their symptom presentation.

Data synthesis and analysis

Sample size calculations were based on data provided by Rundell et al,30 reporting that the mean of the pairwise differences between ECT and EVH is 0.014 percentage points, with an SD of 8.35 percentage points. We estimated that with a drop-out rate of 10%, including 60 athletes would provide a power of 80% at a 5% significance level.

Baseline characteristics are reported as means with 95% CIs, except age which is reported in median (range) for continuous data, while categorical data as the number of cases (n) with percentage (%). Potential differences between groups were analysed using Pearson’s χ2 test for categorical data, and the Mann-Whitney U test for continuous data, as these were skewed.

A diagnostic random effects (DerSimonian and Laird) model with a correction factor of 0.5 (only applied to cells where a 0 was present) was used for the sensitivity and specificity analyses, also presented with 95% CI. Since there is no gold standard BPT to confirm LAD, separate analyses were performed for each test comparing with the others as ‘reference’: (1) standardised ECT, (2) unstandardised ECT, (3) EVH, (4) methacholine BPT. All other statistical analyses were performed with the research platform Ledidi Core.36

Patient and public involvement

Athletes and the public were not involved in the design or conduct of this study.

Equity, diversity and inclusion statement

The study team consisted of junior and senior researchers from different disciplines, balanced in gender and age. The study population included male and female athletes from different sports at amateur and professional level. Potential participants may have been excluded due to geographical distances.

Results

Baseline characteristics

Of 60 athletes (median age 17.5, range 16–28 years, 40% females) with LAD symptoms, 67% represented winter sports, 15% swimmers, 43% reported asthma diagnosis and 37% used inhalation corticosteroids and eventual other asthma medications (table 1). All athletes completed the standardised field-based ECT, EVH and methacholine BPT, while five did not complete the unstandardised field-based sport-specific ECT (figure 1). An average of 1.3 (95% CI 1.2 to 1.4) standardised ECTs and 2.9 (95% CI 2.4 to 3.4) unstandardised ECTs were performed to elicit LAD symptoms, and 10 athletes performed more than 2 ECTs.

Table 1
|
Baseline characteristics of athletes with and without a confirmed diagnosis of lower airway dysfunction (LAD)

At least one positive BPT (ie, standardised ECT, unstandardised ECT, EVH or methacholine BPT) was observed in 68% (n=41) of the athletes. Baseline characteristics are reported in table 1 for athletes with or without at least one positive BPT to confirm the LAD diagnosis. Among 19 athletes (32%) without LAD, 2 athletes presented inspiratory breathing symptoms during EVH, were referred to continuous laryngoscopy during exercise and diagnosed with exercise-induced laryngeal obstruction. One athlete had a history of recurrent chronic bronchitis, was referred to bronchoscopy and diagnosed with tracheobronchomalacia (n=1). The 16 athletes without LAD had no clinical presentation indicating further referral.

Ability to detect LAD

Figure 2 demonstrates the ability of the BPTs to objectively detect LAD. In athletes with confirmed LAD, the standardised or unstandardised ECT was positive in 51% (n=21/41) and 49% (n=20/41) of the cases, respectively, while EVH and methacholine BPT each were positive in 32% (n=13/41). In 59% (n=24/41) of the LAD-confirmed athletes, ECTs were the only positive BPT, while both standardised and unstandardised ECTs were positive in 20% (n=7/35) of these athletes. The mean drop in FEV1 postchallenge for a positive EVH was 16% (95% CI 10% to 22%), for positive standardised ECTs 20% (95% CI 17% to 22%) and positive unstandardised ECTs 15% (95% CI 13% to 17%). Time courses for FEV1 (l) in positive ECTs are demonstrated in figure 3.

Figure 2
Figure 2

Venn diagram illustrating athletes with a positive eucapnic voluntary hyperpnoea test, methacholine bronchial provocation test), standardised or unstandardised field-based exercise challenge test. Closed circles represent athletes with a current asthma diagnosis, and open circles represent athletes with a previous asthma diagnosis.

Figure 3
Figure 3

Time course FEV1 (l) from pre-exercise to postexercise for cases with a positive standardised (left, A, n=21) or unstandardised (right, B, n=20) field-based exercise challenge test. Closed symbols: Athletes with a negative eucapnic voluntary hyperpnoea and methacholine bronchial provocation test. Open symbols: Athletes with a positive eucapnic voluntary hyperpnoea test (diamonds), a positive methacholine bronchial provocation test (triangles) or both tests positive (squares). FEV1, forced expiratory volume in 1 s.

The standardised ECT activity chosen was running 73% (n=44), cross-country skiing 18% (n=7), cycling 7% (n=4) and rowing 2% (n=1). The weather during the standardised ECT was sunny 43% (n=26), cloudy 43% (n=26), foggy 5% (n=3), snowy 5% (n=3) and windy 2% (n=1). The outdoor temperature ranged from −10°C to 10°C in 86% (n=18/21) of the positive standardised ECTs and 74% (n=29/39) of the negative standardised ECTs (p=0.06, χ2 test). For the unstandardised ECTs, weather, temperature, workload and intensity were not reported since they aimed to assess the actual impact on lung function during sport-specific training sessions in their natural environment.

Diagnostic test performance

Tables 2 and 3 present the sensitivity and specificity of the different BPTs. Standardised and unstandardised ECTs confirmed LAD with 54% (95% CI 25% to 81%) sensitivity and 70% (95% CI 55% to 83%) specificity and 46% (95% CI 19% to 75%) sensitivity and 68% (95% CI 53% to 81%) specificity, respectively, with similar results regardless of reference. In comparison, EVH and methacholine BPT were 33% sensitive (95% CI 15% to 57%) and 85% (95% CI 70% to 94%) specific against standardised ECT as a reference.

Table 2
|
Sensitivity analyses (%) for lower airway dysfunction in athletes using each test as the reference standard (value with 95% CI)
Table 3
|
Specificity analyses (%) for lower airway dysfunction in athletes using each test as the reference standard (value with 95% CI)

Discussion

In this study on 60 athletes with LAD symptoms, a positive BPT confirmed the LAD diagnosis in 41 cases (68%). In 24 of these (59%), the novel concept of using an app-based spirometer for unsupervised standardised and/or unstandardised field-based ECTs was the only positive BPT, but with overlap between standardised and unstandardised ECTs in only 20% of the cases.

The 68% prevalence of confirmed LAD diagnosis in symptomatic athletes is higher than in most previous studies,9–11 37 including the 22% prevalence reported in a recent systematic review of LAD in athletes.2 This discrepancy may be explained by the high proportion of winter endurance athletes and pool-based swimmers in our study population, as previous papers including similar populations report up to 70% prevalence of LAD.2 29 34 38

Detecting LAD in symptomatic athletes with unsupervised field-based ECTs

The large proportion of confirmed LAD in our study is mainly explained by the 59% of BPT positive athletes who had a positive ECT only. First, one could question if this finding results from the technology failure of the app-based turbine-spirometer. Degryse et al23 validated supervised manoeuvres on a turbine spirometer with similar technology against a daily calibrated Jäger MasterScope as a reference standard. The correlation between 908 parallel measurements on 34 different patients was good (r2=0.95), and the authors concluded that the Jäger device and the turbine spirometer could be used interchangeably.23 Second, the reliability of unsupervised lung function manoeuvres is secured through the strict built-in ATS/ERS guidelines.20 A study including 7777 spirometries,39 spirometries assessed by computers with built-in guidelines rejected more less valid manoeuvres than human reviewers. The reliability is further supported by (1) the finding of an expected, smooth curve demonstrating at least two consecutive drops in FEV1 of ≥10% postchallenge in all athletes with a positive standardised ECT (figure 3) and (2) the significant correlation observed between positive standardised ECTs and positive EVH. Third, the BPT chosen is important as they differ in test performance. Field-based and sport-specific ECTs are reported as superior to laboratory ECTs,30 and superior to EVH and methacholine BPTs in detecting exercise-induced airway narrowing,14 probably because of the impact of ambient conditions. The limited overlap between the standardised and unstandardised ECTs may result from the less strict criteria for a positive unstandardised ECT as compared with the standardised ECTs (one single vs two consecutive ≥10% drops in FEV1 postchallenge). However, differences in workload and intensity may also play a role, and one may speculate if standardising ECTs can result in a weaker stimulus to elicit LAD. Lack of overlap has been demonstrated between several different BPTs, which emphasises why performing different BPTs is important when LAD is suspected.14 Finally, the app-based spirometer facilitates regular testing of the fluctuating condition of LAD; exercise-induced airway narrowing is often normalised after a few weeks of rest.40 41

Diagnostic test performance

The sensitivity and specificity for documenting LAD with the field-based standardised and unstandardised sport-specific ECTs were 54% and 70%, and 46% and 68%, respectively. Compared with the ECT 51% sensitivity and 84% specificity of ECTs in a recent systematic review on BPT test performance in confirming LAD in athletes,14 our findings demonstrated a similar sensitivity and a somewhat lower specificity. Cool outdoor temperature ranged from −10°C to 10°C for most of the positive standardised ECTs and may have improved ECT sensitivity, as cool air is reported to improve test performance.42

For both EVH and methacholine BPT to document LAD in this study, the sensitivity and specificity were 33% and 85%, respectively, while in a recent systematic review,14 the sensitivity and specificity for EVH were 46% and 74%, respectively, and for methacholine BPT they were 55% and 56%.14 The somewhat poorer sensitivity and higher specificity in this study may result from our strict criteria for a positive EVH; two consecutive drops of ≥10% in FEV1 postchallenge, while several studies require one single drop of ≥10% in FEV1 postchallenge.32 34 43 44 For the methacholine BPT the explanation may be the same based on our strict requirement of PD20<4 µmol for a positive test, while diagnostic cut-off levels range from 4 to 9.47 µmol in previous studies.14 32 45–47

Limitations

A challenge when comparing different BPTs performances to confirm LAD in athletes is that no gold standard exists as a reference. Even though EVH and methacholine BPTs had the highest specificity, ECTs had the highest sensitivity, which is preferable to ensure early detection and allow initiation of appropriate treatment.

Using an app-based spirometer is that technology or network may fail. Also, the method relies on the athletes to ensure that HR is sufficient and that exercise is continuous. Education on lung function performance on the app-based spirometer is particularly important for unsupervised tests.

For the unstandardised ECTs, a strategy with two time points rather than one single drop in FEV1≥10% postchallenge would have increased test reliability. We suggest that if unstandardised ECTs are used to diagnose LAD, several positive tests with consistent results are needed, to reduce the risk of both false negative and false positive tests.

Clinical implications

By including actual ambient conditions, workload and intensity that affect lung function, and by enabling several tests for the fluctuating condition of LAD, this study demonstrates that unsupervised standardised and unstandardised field-based ECTs can support LAD diagnostics in athletes. However, future studies should include supervised ECTs with calibrated spirometers to verify the results. The app-based spirometer is easily available and at low cost, which may reduce the worldwide gap in access to appropriate assessment for LAD-symptomatic athletes.

Conclusion

This study indicates that using an app-based spirometer for unsupervised field-based ECTs in athletes with LAD symptoms can support the diagnostic process for the fluctuating condition LAD.