Article Text

Diagnostic approach to lower airway dysfunction in athletes: a systematic review and meta-analysis by a subgroup of the IOC consensus on ‘acute respiratory illness in the athlete’
  1. Tonje Reier-Nilsen1,2,
  2. Nicola Sewry3,4,
  3. Bruno Chenuel5,6,
  4. Vibeke Backer7,8,
  5. Kjell Larsson9,
  6. Oliver J Price10,11,
  7. Lars Pedersen12,
  8. Valerie Bougault13,
  9. Martin Schwellnus3,4,
  10. James H Hull14,15
  1. 1 The Norwegian Olympic Sports Centre, Norwegian Olympic and Paralympic Committee and Confederation of Sports, Oslo, Norway
  2. 2 Oslo Sports Trauma Research Center, Department of Sports Medicine, Norwegian School of Sport Sciences, Oslo, Norway
  3. 3 Sport, Exercise Medicine and Lifestyle Institute (SEMLI), Faculty of Health Sciences, University of Pretoria, Pretoria, South Africa
  4. 4 International Olympic Committee (IOC) Research Centre of South Africa, University of Pretoria, Pretoria, South Africa
  5. 5 Centre Hospitalier Régional Universitaire de Nancy, Department of Lung function and Exercise Physiology - University Center of Sports Medicine and Adapted Physical Activity, Université de Lorraine, Nancy, France
  6. 6 Université de Lorraine, DevAH, Nancy, France
  7. 7 Department of ENT, Rigshospitalet, Copenhagen University, Copenhagen, Denmark
  8. 8 CFAS, Rigshospitalet, Copenhagen University, Copenhagen, Denmark
  9. 9 Integrative Toxicology, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden
  10. 10 School of Biomedical Sciences, Faculty of Biological Sciences, University of Leeds, Leeds, UK
  11. 11 Leeds Institute of Medical Research at St. James’s, University of Leeds, Leeds, UK
  12. 12 Department of Respiratory Medicine and Infectious Diseases, Bispebjerg Hospital, Copenhagen, Denmark
  13. 13 Laboratoire Motricité Humaine Expertise Sport Santé, Université Côte d’Azur, Nice, France
  14. 14 Department of Respiratory Medicine, Royal Brompton Hospital, London, UK
  15. 15 Institute of Sport, Exercise and Health (ISEH), Division of surgery and Interventional science, University College London, London, UK
  1. Correspondence to Dr Tonje Reier-Nilsen, The Norwegian Olympic Sports Centre, Norwegian Olympic and Paralympic Committee and Confederation of Sports, Oslo 0806, Norway; tonjereiernilsen{at}


Objectives To compare the performance of various diagnostic bronchoprovocation tests (BPT) in the assessment of lower airway dysfunction (LAD) in athletes and inform best clinical practice.

Design Systematic review with sensitivity and specificity meta-analyses.

Data sources PubMed, EBSCOhost and Web of Science (1 January 1990–31 December 2021).

Eligibility criteria Original full-text studies, including athletes/physically active individuals (15–65 years) who underwent assessment for LAD by symptom-based questionnaires/history and/or direct and/or indirect BPTs.

Results In 26 studies containing data for quantitative meta-analyses on BPT diagnostic performance (n=2624 participants; 33% female); 22% had physician diagnosed asthma and 51% reported LAD symptoms. In athletes with symptoms of LAD, eucapnic voluntary hyperpnoea (EVH) and exercise challenge tests (ECTs) confirmed the diagnosis with a 46% sensitivity and 74% specificity, and 51% sensitivity and 84% specificity, respectively, while methacholine BPTs were 55% sensitive and 56% specific. If EVH was the reference standard, the presence of LAD symptoms was 78% sensitive and 45% specific for a positive EVH, while ECTs were 42% sensitive and 82% specific. If ECTs were the reference standard, the presence of LAD symptoms was 80% sensitive and 56% specific for a positive ECT, while EVH demonstrated 65% sensitivity and 65% specificity for a positive ECT.

Conclusion In the assessment of LAD in athletes, EVH and field-based ECTs offer similar and moderate diagnostic test performance. In contrast, methacholine BPTs have lower overall test performance.

PROSPERO registration number CRD42020170915.

  • Asthma
  • Diagnosis
  • Athletes
  • Exercise Test
  • Respiratory System

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • Lower airway dysfunction (LAD) (including exercise-induced asthma and/or exercise-induced bronchoconstriction and/or airway hyperresponsiveness) is highly prevalent affecting approximately one in five athletes.

  • Studies have consistently demonstrated a poor relationship between the presence of respiratory symptoms and objective evidence of LAD in athletes.

  • Bronchial provocation testing is recommended to confirm a diagnosis of LAD, but there is no clear or established ‘gold-standard’ test in this context.


  • There exists a lack of consistency in studies describing the use of bronchial provocation tests in the diagnosis of LAD in athletes, with heterogeneous application of protocols and cut-off values.

  • In athletes reporting symptoms of LAD, both the eucapnic voluntary hyperpnoea (EVH) and field-based sport-specific exercise challenge test had a moderate specificity for the detection of LAD.

  • Field-based sport-specific exercise challenge tests, particularly if performed in a dry environment and at a high intensity/workload, demonstrated greater test performance in comparison to EVH in this context.


  • This study provide an evidence-based guidance for the clinical diagnostic approach to LAD in the athlete.


Lower airway dysfunction (LAD) is a term used to describe asthma-related issues in athletes, including exercise-induced asthma, exercise-induced bronchoconstriction (EIB) and/or airway hyperresponsiveness. These entities collectively represent the most common reason for an athlete to seek medical review.1

In the diagnostic assessment of LAD, several challenges arise. The typical symptoms of LAD (ie, wheeze, cough, chest tightness/dyspnoea and excessive mucus) are non-specific and present in several of the differential diagnoses, including but not limited to exercise-induced laryngeal obstruction (EILO) and breathing pattern disorder.2 This limits the diagnostic precision of a symptom-only/clinician-based approach to the assessment of LAD,2 and prompts the need for objective testing.3 In this context, a broad range of diagnostic tests are frequently used, however, there remains equipoise on the optimal approach. It also remains unclear how test modalities compare with each other and their utility for ‘ruling in’ or ’ruling out’ a diagnosis of LAD, in an athletic population.

Intuitively, the best way to diagnose an exercise-related pulmonary issue would be to assess lung function and specifically the physiology of airflow limitation, before and following a relevant period of exercise, that is, by performing an exercise challenge test (ECT). This approach, however, is challenging because of the need to employ certain standardised work protocols (ie, there is a requirement for a short, very high intensity exercise bout with no preceding warm-up) and to control environmental conditions; all factors that influence the specificity and sensitivity of a subsequent result.4 5 Thus, the athlete must be able to exercise at high intensity (>80% of maximum heart rate) and cannot be injured or recovering from injury, and ECTs may be challenging to schedule, given they will impact on an athlete’s training and competition schedule. Several guidelines have recommended the use of ‘surrogate’ tests for EIB, most often with bronchoprovocation testing (BPT), using inhaled challenge methodologies.4 Indirect BPT, with the eucapnic voluntary hyperpnoea (EVH) or inhaled tests (eg, inhaled mannitol or nebulised adenosine 5′-monophosphate (AMP)) are often cited as representing the ‘gold standard’ for diagnosing EIB, given they act to mimic the desiccating process that promotes EIB in susceptible athletes.3 In contrast, other forms of BPT, such as direct BPTs, act by inhalation of, for example, either methacholine or histamine, to directly stimulate sensitised bronchial smooth muscle and thereby provoke bronchoconstriction independent of inflammation.6 7 In the assessment of asthma in the general population, it has been proposed that indirect BPT are helpful in ‘ruling in’ a diagnosis of asthma when positive, whereas, direct BPT have their highest utility when negative, that is, in terms of ‘ruling out’ a diagnosis.8

In the assessment of LAD in athletes, the sports and exercise medicine clinician is typically faced with a decision on the choice of diagnostic test for LAD in two common clinical scenarios: (1) to confirm a diagnosis of LAD in an athlete presenting with non-specific symptoms of LAD (ie, wheeze, cough, chest tightness/dyspnoea and excessive mucus) and (2) to potentially screen for LAD in athletes, during a periodic health assessment/preseason or precompetition assessments.9 The aim of any approach to diagnostic testing should be to inform the selection of a subsequent treatment plan, that is, enacted to optimise an athlete’s health and ability to undertake exercise without symptoms.10

With these considerations in mind, the aim of this work was to systematically review the available evidence comparing diagnostic test modalities in the context of (1) confirming a symptom-based diagnosis of LAD and (2) screening for LAD in athletes, regardless of symptoms. The study uses a sensitivity and specificity meta-analysis, to inform clinicians regarding the rule in and rule out value of different diagnostic approaches for the diagnosis of LAD. In the absence of a reference diagnostic test or ‘gold standard’, we report and compare the performance of a symptom-based diagnosis against different diagnostic tests modalities.


Protocol and registration

This systematic review and meta-analysis was performed in accordance with the 2020 Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.11 The review was registered prospectively with the PROSPERO database (registration number: CRD42020170915).

Study selection and eligibility criteria

PubMed, EBSCOhost and Web of Science (core collection) databases were used to search for published articles between 1990 and December 2021 using a combination of the terms (eg, asthma OR EIB AND athletes AND screening OR diagnosis) and relevant exclusions. For the full search string for each database, see online supplemental file 1. The results of these searches were combined, and duplicate articles removed. Any additional relevant articles identified by the authors or sourced from the reference list of identified studies were included. All article screening and selection was undertaken using the online tool CADIMA.12

Supplemental material

Inclusion and exclusion criteria

Studies were required to meet the following criteria for inclusion: (1) study participants were athletes/physically active individuals (adult (aged 15–65 years), athletes or physically active individuals at either amateur or professional level); (2) participants had undergone assessment for LAD symptoms by patient recall or questionnaires and/or objective testing (ie, direct or indirect BPT’s) for LAD and (3) original full-text studies (ie, not research correspondence or case studies) of observational, prospective, retrospective, cross-sectional, longitudinal or intervention design, written in English. Animal or non-human studies were excluded. Articles were also excluded if the study was conducted with a heterogeneous sample (ie, mixed sample of athletic and non-athletic populations) without reporting group findings separately, or if it was a review article, expert opinion or consensus position statement. The articles were screened independently by three reviewers in pairs (either TR-N/LP or TR-N/BC) first by title/abstract and then full text, and any conflicts were resolved through discussion to reach consensus.

Data extraction

The data extracted from the included studies are presented in table 1 and divided into four groups: (1) participants (number, age, sex), (2) type of sport and athletic standard, (3) prior physician diagnosed asthma (PDA) and (4) presence of symptoms of LAD (indicating uncontrolled or undiagnosed LAD). Diagnostic methodologies and protocols are presented in table 2. All data were extracted by TR-N and BC and any conflicts resolved through discussion.

Table 1

Summary of study characteristics (n=31)

Table 2

Summary of study methods and protocols (n=31)

Quality assessment and risk of bias

A modified Downs and Black checklist13 was used to determine the quality of the article including a 13-point scale (see online supplemental file 2) for modified version). Two reviewers (TR-N and BC) scored the articles independently and reached consensus on the final score after discussion. The Downs and Black checklist was modified to remove domains pertaining to randomised controlled trials, and included components of reporting (up to seven points), external (up to two points) and internal validity (bias and selection bias) (up to four points) and yielded a final score for each article. The quality assessment score was determined against the following criteria: 11–13: excellent; 9–10: good; 7–8: fair and ≤6: poor. The level of evidence was also determined using the 2009 Oxford Centre for Evidence-Based Medicine Levels of Evidence.14

Supplemental material

Outcome measures

The primary outcome was the sensitivity/specificity of the diagnostic tool to detect LAD. Diagnostic tools included symptoms of LAD and a least one form of BPT. The three ‘reference standards’ for the sensitivity/specificity meta-analysis were: (1) symptoms of LAD, (2) an ECT and (3) EVH. Other studies, where there were a limited number of studies reporting use of a BPT (eg, AMP), or where the authors only compared multiple ECT protocols were not included in the quantitative meta-analyses as they did not have another reference to compare with.

Data synthesis and analysis

A qualitative synthesis of evidence was conducted for all studies. Data are reported as mean±SD or 95% CIs unless otherwise stated. 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 analysis. The 95% CIs for the sensitivity and specificity are also presented, as well as the I2 values (a measure of the heterogeneity of the data). A separate analysis was performed for each ‘reference standard’ (1: symptoms of LAD, 2: an ECT and 3: EVH). An HSROC analysis was not possible due to low numbers (the model did not converge). OpenMetaAnalyst was used for all analyses, a 0.05 level of significance was accepted and data were plotted using PRISMA for visual purposes.


Included studies and quality characteristics

In total, 968 studies were identified. Of these, 31 studies15–45 were included in the qualitative synthesis of study characteristics (table 1); 5 studies were excluded from the quantitative sensitivity and specificity meta-analyses on BPT performance because they did not report all data required for analyses (figure 1). Indirect BPT data were reported in 18 studies and direct BPT results in three studies, while 7 studies reported data on both direct and indirect BPTs. Three studies reported data on symptoms only, with no BPT data included (table 2). Downs and Black Quality Assessment Scores ranged from 10 to 12 and studies were rated as excellent (n=30) or good (n=1) (online supplemental file 3).

Supplemental material

Figure 1

PRISMA flow chart representing search results. PRISMA, Preferred Reporting for Systematic Reviews and Meta-Analyses.

Participant demographics and clinical characteristics

The qualitative synthesis from the included studies, describe a total sample size of n=3083 athletes, with an age range of 15–61 years (table 1). Of the 28 studies that provided full demographic details, 36.0% of the participants were female, while in the 26 studies included for the quantitative meta-analyses on BPT performance (n=2624; 33% female). Winter sport-based athletes were the most common athletic group described (10 studies, n=477 winter athletes), followed by summer sport athletes (9 studies, n=507 various summer sports) and swimmers (7 studies, n=267 swimmers). In the qualitative analyses, the presence of a prior PDA was reported in 25 studies (n=400; 22.0% of the 1811 participants), while a diagnosis of LAD based on the presence of specific respiratory symptoms was reported in n=688 (51.3% of the 1342 participants), detailed in 14 studies. In the quantitative meta-analyses, n=319 (23.6% of 1352 participants) had a prior PDA and n=651 (59.7% of 1311 participants) had a symptom-based LAD diagnosis.

The characteristics and proportion of prior PDA and symptoms of LAD in the included papers are summarised in table 1. The presence of respiratory symptoms was reported in 15 papers.20 21 23–25 27–32 34 37 43 45 These data were obtained using existing or modified standardised questionnaires (Allergy Questionnaire for Athletes)46 in four studies23 28 30 31 and non-validated investigator initiated questionnaires in eight studies20 21 24 25 29 34 42 45; three papers did not report on the use of questionnaires.27 37 43

Diagnostic test protocols reported

A wide variety of test protocols and cut-off levels were reported (table 2). All nine papers including methacholine BPT required ≥20% fall in FEV1 postchallenge and at a specific accumulated provocation dose (PD20), or accumulated provocation concentration (PC20), as per convention. In papers using PD20, the diagnostic cut-off levels for a positive methacholine BPT ranged from 4 µmol to 9.47 µmol,15 27 35 37 38 43 while papers using PC20 had a cut-off of 4 mg/mL methacholine (equivalent to accumulated 8 µmol methacholine).17 18 32

In the 16 papers describing EVH, target ventilation rate was described as 30 × FEV1 (equivalent to 85% of maximum voluntary ventilation,16–18 21 22 26–28 30–33 37 43 44 while test duration time was either 617 18 22 26–28 31–33 44 45 or 8 min.16 21 30 37 The cut-off value for a diagnostic test was a single ≥10% fall in FEV1 postchallenge in the majority of studies,17 18 21 26–28 33 while some papers required a single ≥15% fall in FEV1 postchallenge or a ≥10% fall in FEV1 at two consecutive time points within 30 min postchallenge.16–18 21 30 32 37 44 45

Of the 13 papers including detail on ECTs, these reported various sport-specific field-based ECTs in training or competition ranging in time from few minutes (speed skating races) to several hours (long-distance cross-country or triathlon competition), and in temperatures from +10°C to −15°C.19 22 23 25 28 29 33–37 42 44 One study reported an ECT in a chlorinated pool.44 Three of the ECTs also included an indoor laboratory treadmill test.19 22 42 The cut-off value for a diagnostic test was a ≥10% fall in FEV1 once postchallenge in all,19 22 23 25 28 29 33–37 42 but one paper44 requiring ≥10% reduction in FEV1 at two consecutive time points within 30 min postchallenge.

Diagnostic tests to confirm a symptom-based diagnosis of LAD

Figure 2 includes 12 studies comparing five BPT methodologies (methacholine BPT, AMP, Mannitol, ECT, EVH) with symptoms of LAD as the reference standard. Four of these studies included more than one BPT, allowing 15 cross BPT comparisons are to be included in the meta-analyses.

Figure 2

Bronchoprovocation tests compared with the reference standard of previously reported lower airway dysfunction symptoms. Overall sensitivity: 39.6% (29.5%–50.7%) p=0.066, I2=83.1; overall specificity: 81.6% (68.6%–89.9%) p<0.001, I2=82.7. AMP, adenosine 5'-monophosphate; ECT, exercise challenge test; EVH, eucapnic voluntary hyperpnoea; FN, false negative; FP, false positive; TN, test negative; TP, test positive.

Overall, there was a poor level of agreement between BPT results and presence of LAD symptoms, with large discrepancies in the results from both indirect and direct BPTs to identify athletes with LAD; that is, when symptoms of LAD were taken as the reference, the overall sensitivity of BPTs to identify athletes with LAD was 40% (95% CI 30% to 51%). In contrast, the agreement between a negative BPT result and absence of respiratory symptoms was more consistent with a specificity of 82% (95% CI 69% to 90%). Studies evaluating EVH demonstrated an overall 46% sensitivity (95% CI 32% to 62%) and 74% specificity (95% CI 54% to 88%) for a symptom-based LAD diagnosis. The ECTs included in the meta-analyses were all field-based and sport-specific (figure 2 and table 2) and demonstrated a similar specificity (84%, 95% CI 34% to 98%) due to similar means and wide variance, with a somewhat higher sensitivity (51%, 95% CI 39% to 62%). Figure 2 demonstrates that the ECTs performed in colder weather, higher altitudes and higher intensities, demonstrated the highest sensitivities for the LAD diagnosis.23 34 Studies evaluating methacholine BPT demonstrated a lower overall test performance for a symptom-based diagnosis, with a 55% (95% CI 21% to 85%) and 56% (95% CI 40% to 71%) sensitivity and specificity, respectively.

Diagnostic tests to detect LAD regardless of symptoms in athletes, that is, screening for LAD

EVH as the reference standard

Figure 3 details findings from eleven studies comparing respiratory symptoms and BPT methodologies, with EVH as the reference standard. Athletes with symptoms of LAD demonstrated 78% sensitivity (95% CI 57% to 90%) and 45% specificity (95% CI 26% to 66%) for a positive EVH, while a positive ECT was 42% sensitive (95% CI 27% to 59%) and 82% specific (95% CI 66% to 91%) for a positive EVH.

Figure 3

Symptoms and bronchoprovocation tests compared with the reference standard of an eucapnic voluntary hyperpnoea test. Overall sensitivity: 68.0% (50.8%–81.4%) p=0.041, I2=74.3; overall specificity: 63.6% (46.4%–77.9%) p=0.118, I2=85.3. ECT, exercise challenge test; FN, false negative; FP, false positive; TN, test negative; TP, test positive.

ECT as the reference standard

Figure 4 includes ten comparisons in the meta-analyses comparing respiratory symptoms and BPT methodologies with ECT as the reference standard (n=9 studies; 1 multiple comparisons). Athletes with symptoms of LAD demonstrated 80% sensitivity (95% CI 38% to 96%) and 56% (95% CI 39% to 71%) specificity for a positive ECT, while a positive EVH was 65% sensitive (95% CI 34% to 87%) and 65% specific (95% CI –47% to 79%) for a positive ECT.

Figure 4

Symptoms and bronchoprovocation tests compared with the reference standard of an exercise challenge test. Overall sensitivity: 59.8% (40.4%–76.5%) p=0.324, I2=54.5; overall specificity: 62.9% (52.3%–72.4%) p=0.018, I2=64.0. EVH, eucapnic voluntary hyperpnoea; FN, false negative; FP, false positive; TN, test negative; TP, test positive.


In this systematic review and meta-analysis, we evaluated studies conducted over the past 30 years that characterise the diagnostic techniques and approaches used in the assessment of LAD in athletes. Our qualitative analyses included 31 studies that describe diagnostic assessments of LAD in approximately 3000 athletes. Out of these, 26 studies included sufficient data to perform a quantitative meta-analysis, comparing diagnostic test modalities for the diagnosis of LAD in approximately 2500 athletes. This analysis revealed that there is a heterogeneous approach in both the test protocols and diagnostic cut-off values employed, however, the key findings from our analysis indicate that: (1) when the aim for a sport and exercise medicine clinician is to confirm a diagnosis based on the presence of LAD symptoms, EVH and ECTs demonstrated moderate and similar specificity for a diagnosis, given the similar overall mean and variance estimates, (2) in athletes with symptoms of LAD, methacholine BPT demonstrated a lower overall test performance for the diagnosis of LAD and (3) when screening athletes regardless of the presence of symptoms of LAD, a field-based sport-specific ECT at a high intensity level performed in a cold environment may be more sensitive than EVH in the assessment of LAD.

A key difficulty encountered when comparing BPTs to confirm LAD in athletes is determining what should be considered the gold standard or comparator test. In the context of other diagnostic tests, an assessment algorithm employing tests with the highest sensitivity is preferable, to ensure early detection and thus initiation of appropriate treatment. However, the negative effects of misclassification need to be considered in any diagnostic appraisal, including the potential long-term side effects and risks of prescribing unnecessary medications, including associated healthcare costs and psychological implications.47 Hence, clinicians need to be aware of the sensitivity and specificity of any given diagnostic test, in order to successfully apply this test when evaluating symptoms.48 In the diagnostic assessment of athletes with respiratory problems, ultimately the diagnosis will arise following a synthesis between the presence of compatible symptoms and diagnostic tests that support the evidence of LAD. The likelihood of a correct diagnosis is increased if the pretest probability of a diagnosis is high, and the test performance is high quality.48 In contrast, if the pretest probability is low, test result needs to be interpreted with caution.48

Diagnostic tests to confirm a symptom-based diagnosis of LAD

In the meta-analyses, we show that both the EVH and field-based ECTs have reasonable test performance characteristics, if viewed from the perspective of a sport and exercise medicine clinician trying to confirm a diagnosis of LAD in a symptomatic athlete. The overall value of EVH as a diagnostic test for LAD in athletes is that it mimics the pathophysiology of LAD.49 The key pathophysiological mechanism behind LAD is hyperpnoea-induced evaporative water loss from the airway surface, which induces release of local mediators with subsequent bronchoconstriction.50 51 The process is amplified by dry and cold air, air pollution52 and inhalation allergies.51 Hence, the EVH is a highly potent stimulus as it involves a high ventilation rate of dry gas mixture, reported to result in a low false-negative rate for the diagnosis of EIB.49 However, our meta-analysis demonstrated EVH to have an overall moderate specificity (74%) and low sensitivity (46%) to confirm a symptom-based LAD-diagnosis. In contrast to the moderate specificity, Levai et al demonstrated an overall low specificity, which may be impacted by the very high prevalence of LAD in the elite aquatic population and the nature of the study design.45 The low sensitivity of EVH in our meta-analysis was influenced by one study by Parsons et al,31 which may have been influenced by the inclusion of mostly soccer/lacrosse sports at a subelite level, and thus associated with a lower prevalence of LAD.1 53 However, most studies in this meta-analysis show that EVH has a moderate sensitivity. These findings aligns with previous guideline documents, recommending EVH as the ‘gold standard’ to identify EIB in athletes.10 54 55 There does remain some debate regarding the test–retest repeatability, with some studies reporting relatively poor short-term repeatability, especially in the context of mild or borderline EIB,54 56 while others have shown better short-term and long-term test–retest validity for EVH.57 58

A moderate specificity of EVH may increase the false-positive rate. In a study on n=224 asymptomatic athletes,53 as many as 20% had a positive EVH when >10% fall in FEV1 postchallenge was employed as the diagnostic cut-off. The fall in FEV1 following EVH was also more pronounced in elite level, when compared with recreational athletes. To reduce the risk of false positives, some researchers have suggested use of a more conservative cut-off value (eg, >15% fall in FEV1).4 However, one cannot rule out the possibility that athletes whom are reportedly asymptomatic, may be misattributing the perception of dyspnoea as their normal exercise response.59–61 In our meta-analyses, variations in EVH results may also have been influenced both by different diagnostic cut-off values and variations in test duration time.

Our findings indicate that field-based sport-specific ECTs also demonstrate a moderate performance in the confirmation of a symptom-based LAD in athletes, in line with previously published studies.22 59 62 Standardisation is the main challenge with sport-specific ECTs in this context. Reports indicate that exercise load and intensity during sport-specific ECTs most certainly have an impact on occurrence of EIB.62 Furthermore, ambient conditions also have a great impact increasing both specificity and sensitivity of sport-specific ECTs if performed in cold weather23 34 or in chlorinated swimming pools,20 44 whereas humid air may blunt propensity to development of EIB.63 In this meta-analysis, we found that studies reporting ECT in colder ambient temperatures appear to report a higher sensitivity result,19 23 25 28 33 which is in line with previous reports demonstrating that inspiring cold dry air enhances the risk of EIB and inversely, inhaling warm humid air is a weaker stimulus and may even prevent EIB.4 5 64 The sensitivity of an ECT would also be increased by reducing the cut-off of the % fall in FEV1 from the baseline value to diagnose EIB, from 10% to 6.5% as some authors suggested.25

In this meta-analysis, direct BPTs, mainly methacholine BPT, appear to be less specific compared with indirect BPTs in confirming a diagnosis of LAD. This finding is in line with a previous study that reported low sensitivity (<40%) and a low negative predictive value of methacholine BPT in the elite athlete.27 This indicates that the methacholine BPT is a less favourable diagnostic test for athletes in the workup of LAD compared with indirect BPTs.65 Our meta-analysis highlights the variability of the techniques and devices used worldwide to deliver methacholine, making it difficult to be precise about dose/concentration equivalents, especially when concentration is mostly cited in guidelines.66 The question of the cut-off values of methacholine responsible for 20% fall in FEV1 used to diagnose airway obstruction is also crucial. In this meta-analysis, we note that cut-off levels for a positive PD20 varied from 4 µmol to 9.47 µmol administered cumulated methacholine,27 35 clearly contributing to the results. Furthermore, one study also performed a methacholine BPT after an EVH on the same day.18 Performing two bronchial provocation challenges on the same day is not advised, since the first challenge may affect the outcome of the latter, resulting in a possible false positive test.67

To summarise, we have found in this meta-analysis, that EVH and ECTs have similar specificity and are the most precise methods for confirming a symptom-based diagnosis, while ECTs may be more sensitive than EVH. In contrast, direct BPT are less specific in this setting. The same appears to be true in the workup of LAD in the general population.8

Diagnostic tests to detect LAD regardless of symptoms in athletes, that is, screening for LAD

The second aim of this systematic review and meta-analysis was to compare the performance of different BPTs in a ‘screening-type’ context, to inform decision making. When screening, we search for both a high sensitivity and a high specificity. In our meta-analysis, EVH and ECTs demonstrated a similar and moderate specificity due to similar means and a wide range. Regarding sensitivity, however, the meta-analysis demonstrated that ECTs were more sensitive than EVH, particularly if performed in colder weather, higher altitudes and higher intensities.

Hence, for field-based sport-specific ECTs performed at colder temperatures, there was a high agreement between the presence of LAD symptoms and a positive test. These findings are in contrast to included studies reporting no association between respiratory symptoms and a positive BPT.4 23 27 28 30 32 34 68 The discrepancy between symptoms and BPT results may be explained by (1) athletes under-reporting symptoms since respiratory symptoms are expected to increase with exercise load,59–61 (2) coexisting conditions that mimic LAD, such as EILO,2 (3) long time interval between presence of LAD symptoms and the BPT, as BPT results can normalise in some cases after a few week’s rest17 25 and (4) use of direct BPTs rather than indirect BPTs in the assessment of LAD in athletes, which in this systematic review has been well demonstrated may lead to underdiagnosis.

To summarise, EVH and field-based sport-specific ECTs demonstrate a similar and moderate specificity, even though only EVH were significantly specific in detecting LAD by screening athletes regardless of symptoms. However, ECTs may be more sensitive than EVH in this context.

Methodological considerations and future research

Several methodological limitations were evident from the systematic review process. First, there is marked heterogeneity between studies with differences in BPT methods, protocols, sport-types and clinical definitions of LAD and asthma, limiting the ability to make direct and conclusive comparisons between studies. Additionally, the minimal data we have at present may also have resulted in the wide variance in CIs with imprecise results. These factors highlight the need for caution when interpreting the overall mean values from the synthesis and meta-analysis of data sources, and may be the reason why a similar precision for ECTs and EVH were observed. It also highlights a need for future work in this field to adhere to recognised protocols, based on international guidelines for the performance and interpretation of BPT.4 67 Second, it would be preferable with more high-quality data describing other phenotypic features and/or a robust characterisation of asthma-related morbidity (eg, exacerbations or marker of type 2 inflammation such as blood eosinophils and/or fractional exhaled nitric oxide). In the general population, this level of detail is now recognised as central in a process that informs best management. Third, the included studies largely include male subjects and from centres in the USA and Northern Europe and thus, there remains very limited insight from more diverse geographical regions and/or low-resource countries. And finally, restricting the literature search to English language, may also have resulted in a selection bias.


In conclusion, the best available data indicate that EVH and ECTs had similar test performance due to their wide CIs in the diagnostic assessment of LAD in symptomatic athletes. A field-based sport-specific ECT performed in a dry air environment appears to be more sensitive than EVH. In contrast, direct BPTs appear to have lower test performance characteristics for the diagnosis of LAD in athletes. Future work should focus on improved overall characterisation of LAD in athletes with comparison to other features, such as airway inflammation and in different sporting types. Studies should also include data from low-resource countries, to provide a globally inclusive perspective concerning the best way to assess and diagnose LAD in athletes.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.

Ethics statements

Patient consent for publication


Supplementary materials


  • Twitter @ReierTonje, @oliverjprice, @VBougault

  • Contributors Conception and design: TR-N, NS, BC, VB, KL, OJP, LP, VB, MS and JHH. Analysis and interpretation: TR-N and NS. Drafting the manuscript for important intellectual content: TR-N, NS, BC, VB, KL, OJP, LP, VB, MS and JHH. TR-N confirms full responsibility for the content of the manuscript.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • © Author(s) (or their employer(s)) 2023. No commercial re-use. See rights and permissions. Published by BMJ.

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.