Exhaled and Systemic Biomarkers to Aid the Diagnosis of Bronchial Asthma in Elite Water Sports Athletes

ABSTRACT Purpose Our aim was to evaluate the accuracy of a combined airway inflammatory biomarker assessment in diagnosing asthma in elite water sports athletes. Methods Members of the Hungarian Olympic and Junior Swim Team and elite athletes from other aquatic disciplines were assessed for asthma by objective lung function measurements, and blood eosinophil count (BEC), serum total immunoglobulin E (IgE), fractional exhaled nitric oxide (FENO) measurements, and skin prick testing were performed. A scoring system from BEC, FENO, serum IgE, and skin test positivity was constructed by dichotomizing the variables and assigning a score of 1 if the variable is elevated. These scores were summed to produce a final composite score ranging from 0 to 4. Results A total of 48 participants were enrolled (age 21 ± 4 yr, 42% male), of which 22 were diagnosed with asthma. Serum total IgE and FENO levels were higher in asthmatic individuals (68 [27–176] vs 24 [1–43], P = 0.01; 20 [17–26] vs 15 [11–22], P = 0.02), and positive prick test was also more frequent (55% vs 8%, P < 0.01). Asthmatic participants had higher composite variable scores (2 [1–3] vs 1 [0–1], P = 0.02). Receiver operating characteristic analysis showed that total IgE, FENO, and composite variable were suitable for identifying asthmatic participants (area under the curve = 0.72, P = 0.01; 0.70, P = 0.02, and 0.69, P = 0.03). A composite score of >2 reached a specificity of 96.2%, a sensitivity of 36.4%, and a likelihood ratio of 9.5. Logistic regression model revealed a strong association between the composite variable and the asthma diagnosis (OR = 2.71, 95% confidence interval = 1.17–6.23, P = 0.02). Conclusions Our data highlight the diagnostic value of combined assessment of Th2-type inflammation in elite water sports athletes. The proposed scoring system may be helpful in ruling in asthma in this population upon clinical suspicion.


B
ronchial asthma is a common respiratory disease affecting more than 330 million people worldwide (1).Exercise-induced bronchoconstriction (EIB), which is characterized by acute onset of bronchospasm during or immediately after physical exertion, is a common feature of asthma (2).However, it can occur in nonasthmatic individuals as well (3,4), and its exact prevalence varies according to the criteria used for the diagnosis (2).Nonetheless, asthma and EIB are very common among elite athletes and may potentially limit the athletic performance (5)(6)(7)(8)(9)(10).A recent systematic review and meta-analysis showed that the overall prevalence of lower airway dysfunction (a collective term for asthma, EIB, and airway hyperresponsiveness) is more than 20% among athletes; however, in elite athletes, it reaches almost 30%, whereas in aquatic disciplines, it is almost 40% (11).
Diagnostic tools for chronic inflammatory airway diseases involve basic (e.g., questionnaires, full blood count measurement, and spirometry with or without reversibility testing) and specific investigations (peak expiratory flow [PEF] variability measurement, direct (acting on airway smooth muscles, e.g., metacholine challenge) and indirect (acting on airway inflammatory cells, e.g., eucapnic voluntary hyperpnea) bronchial provocation tests, skin prick test, immunoglobulin E (IgE) measurement, etc.).The diagnosis of asthma is very challenging and requires specific considerations in elite aquatic athletes because they have often higher than normal Address for correspondence: Veronika Müller, M.D., Ph.D., D.Sc., Department of Pulmonology, Semmelweis University, Tomo str.25-29, 1083 Budapest, Hungary; E-mail: muller.veronika@med.semmelweis-univ.hu.Submitted for publication June 2023.Accepted for publication February 2024.lung function parameters, the symptoms are mostly present only during exertion and exercise in water, and the symptoms do not necessarily correlate with objective findings (5,12).Therefore, although clinical history and subjective symptoms are important in the diagnostic process, according to the recommendation of the International Olympic Committee, an objective test such as bronchial challenge should be performed (13).Nevertheless, the accessibility to such tests varies across regions and countries.However, certain exhaled and systemic biomarkers associated with asthma-specific and T helper (Th) 2-type inflammation can aid in the diagnosis (14).Blood eosinophil count (BEC), total IgE concentration, and skin prick test results are associated with airway allergy and Th2-type inflammation and may be used in phenotyping asthma (14).Elevated fractional exhaled nitric oxide (F ENO ) is a marker of airway eosinophilic inflammation and a recommended diagnostic tool for bronchial asthma in the normal population (14,15).However, F ENO levels in athletes may be different than of the general population, as highlighted by a recent study where 45% of the participants with EIB had values <25 ppb (16).
Consequently, we aimed to assess the clinical use of a complex biomarker assessment (cutaneous allergy testing, F ENO measurement, BEC, and total IgE) for diagnosing asthma in a cohort of Hungarian elite water sports athletes.

METHODS
Participants.We included aquatic athletes who came for asthma screening as part of a sports medical assessment with-out symptoms, or came for the screening due to respiratory symptoms (such as cough, wheezing, breathlessness, and/or a sudden decrease in training performance) or asthma therapy optimization.In the frame of detailed sports medical screening, all members of the Hungarian Swim Team, participating in the Tokyo Olympics 2021, were screened for asthma (N = 41).We also performed the same screening for the National Junior Swim Team (n = 21) before the European Aquatic Championship 2022.Furthermore, we included symptomatic elite aquatic athletes from other disciplines (non-Olympic swimmers n = 8, water polo n = 8, Paralympic swimming n = 3).
Exclusion criteria included previously diagnosed and treated bronchial asthma and a diagnosis of exercise-induced breathing disorders of different etiologies (e.g., exercise-induced laryngeal obstruction (17)).We also excluded participants for whom the diagnosis of asthma/EIB could not be ruled out or confirmed (e.g., atypical symptoms, indeterminate or borderline elevated biomarker levels, partial reversibility of airflow limitation [<12% but >200 mL]) at the end of the data collection period.In those participants, a longitudinal assessment was performed, which exceeded the timescale of our study.The participant selection process is illustrated in Figure 1.
Design.This was a cross-sectional observational clinical study carried out at the Department of Pulmonology, Semmelweis University, Budapest, Hungary, between March 2021 and June 2022.
First, a thorough clinical assessment was performed, including assessment of clinical history, standard physical examination, the nature of the respiratory symptoms, and their relationship with exercise.Furthermore, participants underwent baseline investigations in the following sequence: venous blood sampling, F ENO measurement, spirometry and body plethysmography with reversibility testing, and skin prick test.Further investigations (e.g., bronchoprovocation testing and peak expiratory flow [PEF] monitoring) were performed if the baseline spirometry and reversibility results were not suggestive of airway obstruction, but the participant had a history of exercise-induced respiratory symptoms and was clinically well at the time.Based on thorough clinical history, clinical signs and symptoms, results of the investigations, and a robust evaluation of the likelihood of the diagnosis, the diagnosis of asthma/EIB was confirmed by a highly trained senior respiratory specialist with experience in sports medicine (5,14,18,19).
The clinical characteristics and results of the investigations were compared between the asthmatic and the nonasthmatic participants.The diagnostic accuracy of BEC, F ENO , total IgE level, and prick test positivity was assessed.BEC, F ENO , and IgE concentrations were evaluated both as continuous and as dichotomized variables with the following predefined cutoff values: BEC, ≥150 cells per microliter; total IgE concentration, ≥100 IU•mL −1 ; and F ENO , ≥25 ppb (20)(21)(22)(23).We developed a scoring system to predict asthma likelihood using dichotomous variables.Each variable was defined as either normal or increased based on the above cutoffs, and we assigned a score of 1 if the variable is elevated and 0 if it is not.These scores were summed to produce a final score ranging from 0 to 4.
The study was designed in accordance with the 1964 Declaration of Helsinki and its later amendments, and the protocol was approved by the ethics committee of the Semmelweis University Regional and Institutional Committee of Science and Research Ethics (approval no.55263-6/2020/EÜIG).
Methods of measurements.Venous blood samples were collected from all participants to measure the full blood count, including eosinophil granulocyte counts and percentages, and serum total IgE concentrations (Sysmex XN-1000, Sysmex Corporation, Kobe, Japan, and Beckman Coulter AU680, Beckman Coulter Inc., Indianapolis, IN).
Skin prick test was conducted according to international guidelines (24,25) by a trained healthcare professional.Participants taking oral antihistamine medications were instructed to stop the medication 5 d before the test.Histamine dihydrochloride served as a positive control, and physiological saline solution was used as a negative control.The tests were interpreted 15-20 min after application, and a positive result was defined as a wheal ≥3 mm in diameter.The following inhaled allergens were tested: Dermatophagoides pteronyssinus, Dermatophagoides farinae, Cladosporium herbarum, Aspergillus fumigatus, Alternaria alternata, dog dander, cat dander, mixed feathers, mixed trees (early and late blossom), grass mix, weed mix, cereal mix, mugwort, and common ragweed (Inmunotek Standard Prick Test Panel for Inhalants, Spiromed Ltd., Budapest, Hungary).
F ENO was measured by a trained specialist using a chemiluminescent NO analyzer (Sievers NOA-280i, GE Analytical Instruments, Boulder, CO) according to international guidelines (21), and as previously described (26).All participants were nonsmokers and were asked to refrain from consuming food or beverages, taking inhaled medications, or exercising for 1 h before measurement.Exhaled NO analysis was performed for all participants before spirometry.The ambient NO level at the time of testing was <5 ppb.During the measurement, the participants were instructed to inhale to neartotal lung capacity (TLC) and then exhale in the device against a calibrated resistance with a constant flow rate of 50 mL•s −1 until a plateau was reached at the NO level for at least 3 s.
Lung function parameters were measured by an electronic spirometer and a body plethysmograph (PDD-301/s, Piston, Budapest, Hungary) according to the European Respiratory Society (ERS) and American Thoracic Society (ATS) guidelines (27).The registered parameters included forced vital capacity (FVC), forced expiratory volume in the first second of exhalation (FEV 1 ), FEV 1 and FVC ratio, forced expiratory flow between 25% and 75% of FVC (FEF 25-75 ), TLC, residual volume (RV), RV and TLC ratio, and airway resistance (R aw ).Three technically acceptable maneuvers were performed with a between-maneuver difference in FVC and FEV 1 ≤ 150 mL, and the highest result was used.The diffusing capacity of carbon monoxide (CO) (DL CO ) and lung volume-corrected diffusing capacity (KL CO ) was measured using the singlebreath method (PDD-301/s, Piston).Pulmonary function variables were expressed as percentages of the predicted values using the Global Lung Function Initiative reference equations (28).In cases of suspected asthma, the following tests were performed according to the Global Initiative for Asthma document and international guidelines: reversibility testing after 400 μg of salbutamol inhalation (positive if FEV 1 increases >200 mL and >12% of the predicted value), indirect bronchial challenge test with hypertonic potassium chloride solution inhalation (positive if FEV 1 falls >15% of the predicted value), and/or daily and pre-and postexercise PEF measurements (positive if variability is >20% over a 2-wk period) (5,14,29).
Statistical analysis.TIBCO Statistica (version 13; TIBCO Software Inc., Palo Alto, CA) and IBM SPSS (version 28; IBM Corp., Armonk, NY) statistical software packages were used for data analysis.Continuous variables were compared using unpaired t-tests or Mann-Whitney U-tests according to the distribution of variables.Categorical variables were analyzed using the chi-squared test or Fisher exact test.Receiver operating characteristic (ROC) curves were generated, and the area under the curve (AUC) was calculated to assess the diagnostic accuracy of F ENO , BEC, total IgE measurements, skin prick test, and the composite variable generated from these variables after dichotomization.We used Youden's index to determine the optimal cutoff values based on ROC analysis.The index is determined by calculating the point on the curve where the sum of sensitivity and 1 specificity is the highest (30).Furthermore, binomial logistic regression models were used to explore the variables that predicted the diagnosis of asthma.

RESULTS
The detailed clinical and demographic characteristics of the participants are summarized in Table 1.A total of 48 participants were enrolled in the study, of which 22 (46%) were diagnosed with asthma at the end of the clinical investigation.There was no difference in age or sex distribution between the asthmatic and the nonasthmatic groups.Participants in the asthmatic group had lower FEV 1 % of the predicted value, FEV 1 /FVC ratio, FEF 25-75 values, and increased R aw compared with nonasthmatic participants.The diffusing capacity of asthmatic participants was lower than that of nonasthmatics.Furthermore, the change in FEV 1 , both in raw value and in percentage of the predicted value, was greater in the asthmatic group during reversibility testing.Forty-two participants had a reversibility testing, of which 6 (14%) showed a positive response.Ten participants had bronchial challenge, of which 7 (70%) exhibited a positive response.
The ROC curves of the individual biomarkers are shown in Figure 3. BEC and eosinophil granulocyte percentages had an area under the ROC curve of 0.51 (95% confidence interval [CI] = 0.34-0.68)and 0.50 (95% CI = 0.33-0.67),respectively.Total IgE and F ENO concentrations were suitable for discriminating asthmatic participants from nonasthmatic individuals with AUC values of 0.72 (95% CI = 0.57-0.86,P = 0.01) and 0.70 (95% CI = 0.55-0.85,P = 0.02).We calculated the optimal cutoff values of IgE and F ENO based on Youden's index to balance sensitivity and specificity.An IgE cutoff of 30 IU•mL −1 yielded a sensitivity of 77% and a specificity of 65%.A F ENO cutoff of 20 ppb had a sensitivity of 72% and specificity of 65%.
We created a composite variable from the BEC, total IgE, F ENO , and prick test results.BEC, total IgE, and F ENO levels were dichotomized with cutoff values of ≥150 cells per microliter, ≥100 IU•mL −1 , and ≥25 ppb, respectively.Each component was given a score of 0/1, and the sum of the scores was used as an ordinal variable.Asthmatic participants had significantly higher scores than nonasthmatics (2 [1-3] vs 1 [0-1], Mann-Whitney U-test P = 0.02).ROC analysis of the scoring system was performed.The composite biomarker variable could distinguish between the asthmatic and the nonasthmatic participants, with an AUC of 0.69 (95% CI = 0.53-0.85,P = 0.03).The best cutoff value to balance sensitivity and specificity was 2 points, with a sensitivity of 54.5%, a specificity of 80.8%, a positive predictive value (PPV) of 70.6%, a negative predictive value (NPV) of 67.7%, and a likelihood ratio of 2.8.However, a cutoff point of 3 achieved a specificity of 96.2%, a PPV of 88.9%, and a likelihood ratio of 9.5, meaning that a result of >2 points gives a very strong prediction of asthma in this population (Table 2).However, it is important to note that the sensitivity at this cutoff remains relatively low at 36.4%, with an NPV of 64.1%.
Binary logistic regression was performed to assess the effects of sex, weight, baseline FEV 1 %, and the composite biomarker variable on the likelihood of having asthma (Table 3).The regression model was statistically significant (chi-square = 14.5, P = 0.04), explained 52.4% of the variance in asthma diagnosis (Nagelkerke R 2 = 0.54), and correctly classified 84.8% of cases.Each score increment in the composite variable increased the likelihood of asthma by 2.7 times.Increased weight was associated with an increased likelihood of asthma.The baseline FEV 1 % was negatively associated with asthma.Sex did not

BASIC SCIENCES
significantly contribute to the model.The individual components of the composite variable were not tested in the model because they are not independent (i.e., they share a common pathophysiological background), and it would violate the assumptions of the binary logistic regression.Age was excluded from the model because the age of the participants was homogenous and all were young adults; therefore, it had no clinical relevance.

DISCUSSION
To the best of our knowledge, this is the first clinical study designed to evaluate a complex biomarker assessment for diagnosing bronchial asthma in a cohort of elite water sports athletes.We have proposed an easy-to-use scoring system based on Th2-type inflammation markers for elite water sports athletes with respiratory symptoms, which is strongly associated with a positive diagnosis of asthma and may therefore aid in the diagnosis.However, it is essential to acknowledge the limitation of our proposed diagnostic method, i.e., its low sensitivity and negative predictive value.Furthermore, we have confirmed the previously recognized high prevalence of asthma within this population.
Lower airway dysfunction is common in the general population and are more frequent in elite athletes.Levai et al. (31) found that the prevalence of EIB measured by indirect bronchial provocation tests can reach almost 70% in elite swimmers.Meanwhile, a large-scale international study conducted by collecting therapeutic use exemptions data of elite competitive athletes found that the combined prevalence of asthma and airway hyperresponsiveness is around 22% in aquatic

BASIC SCIENCES
disciplines (7).The huge discrepancies in reported prevalence may be explained by the different study designs (performing new measurements vs collecting existing data), sample sizes, and assessed population (national vs international).Nevertheless, a recent systematic review and meta-analysis summarizing the results of studies involving more than 37,000 participants concluded the prevalence of lower airway disfunction in water sports athletes lies around 40% (11).The reason for the high prevalence of lower airway pathologies in aquatic sports is mainly thought to be caused by the chronic inhalation of volatile chlorine by-products, e.g., trichloramines originating from the disinfection of water pools (32) and by repeated airway epithelial microinjuries caused by sheer stress due to high minute ventilation (33).However, numerous studies have attempted to link chlorine products to the increased prevalence of asthma among swimmers, with conflicting results (32,34,35).Bougault et al. (36) collected respiratory tissue samples from nonasthmatic elite swimmers, asthmatic nonswimmers, and nonasthmatic nonathletes.They found that swimmers had inflammatory and remodeling changes similar to asthmatics and concluded that this was due to long-term exposure to chlorinated swimming pools.On the other hand, Llana-Belloch et al. (34) did not find markers of increased oxidative stress or lung damage after 40 min of exercise in indoor swimming pools despite significant exposure to chlorine by-products.Nevertheless, it is likely that chlorine by-products, mainly trichloramines, act as sensitizers against airborne allergens, leading to the development of atopy and airway remodeling (7,31,(36)(37)(38).However, the increased prevalence of exercise-induced airway pathologies in swimmers is likely due to a complex mechanism involving genetic and epigenetic predisposition, airway inflammation and remodeling caused by recurrent viral respiratory infections, and chronic epithelial damage and sensitization from exposure to chlorine by-products (32,39).Regardless of the underlying cause, asthma-specific, mainly Th2-driven, inflammation is present in this population as well, and certain markers of Th2 inflammation may aid in the diagnosis and monitoring of the disease.
However, is it paramount to acknowledge that more than 60% of the newly diagnosed asthmatic individuals did not have Th2-type inflammation as assessed by our composite variable, and it limits the applicability of our proposed scoring system.Indeed, Couto et al. (40) have proven by latent class analysis that there are two types of asthma in athletes, one is the classical "atopic asthma," driven by Th2-type inflammation, and the other is a non-Th2-type disease, which the authors call "sports asthma," defined by exercise-induced airway symptoms without features of atopic allergy.Importantly, sports asthma is not associated with elevated F ENO or other classic markers of asthma and is more common in aquatic and winter sports.Couto et al. have found that participation in competitive water sports was associated with an almost threefold increase in the odds of having this type of disease, whereas winter sports athletes had an odds ratio of more than 8.5 compared with other athletes.Several mechanisms have been proposed that may lead to the development of sports asthma (33,41).For example, dehydration due to thermal changes and increased ventilation and consequent osmotic changes may lead to bronchoconstriction via bronchial wall edema and airway smooth muscle contraction through the activation of inflammatory pathways and changes in airway vascular diameter (33,41,42).However, dehydration seems unlikely in water sports athletes considering the humid environment above the water surface.In this population, the microtrauma theory is more likely, suggesting that athletes suffer repeated airway epithelial injuries due to the mechanical stress caused by high minute ventilation, exposure to chemicals, and microaspirations of water droplets (43).The repeated injury and healing process leads to altered contractile function of the small airways, contributing to the development of exercise-induced airway symptoms (44).This theory might be further supported by the observation that the inflammatory and remodeling processes may be reversed in some cases by quitting swimming (45).Nevertheless, the high prevalence of "sports asthma" most likely contributed to the low sensitivity and negative predictive value of our composite diagnostic test based on Th2-type biomarkers.
Nitric oxide is a gaseous molecule that can be measured in the exhaled breath (21).The fraction of exhaled NO (F ENO ) is associated with eosinophilic airway inflammation and is therefore an important biomarker of bronchial asthma (14).However, significant intra-and interindividual variation can be observed in the levels of F ENO , which necessitates the setup and use of a standardized and reliable measurement protocol, as endorsed by the ATS and the ERS (21).Accordingly, based on the high variability and the observation that F ENO can be elevated in conditions other than asthma, the Global Initiative for Asthma document does not recommend the use of F ENO for ruling in or ruling out the diagnosis of asthma (29).Nevertheless, the diagnostic usefulness of F ENO is recognized by the ERS in their most recent asthma diagnosis guideline and by the clinical statement of the British Thoracic Society (BTS) addressing the diagnosis, evaluation, and management of respiratory problems in athletic individuals (5,14).The ERS recommends a cutoff of >50 ppb for a specificity of >90% or a cutoff of >40 ppb for the best compromise between sensitivity and specificity.In support of this, Dickinson et al. (16) have reported a study involving recreational and elite athletes from a wide range of athletic disciplines, that F ENO ≥ 40 ppb predicts EIB with the best specificity/sensitivity ratio (86% and 37%, respectively).However, given the low sensitivity, the authors recommend against using F ENO alone in diagnosing EIB.Similarly, the ERS guideline emphasizes that all biomarkers, including F ENO , must be interpreted in the context of the symptoms and the clinical picture, and a low F ENO level does not rule out asthma.The BTS recommends measuring F ENO as part of the diagnostic workup in athletes, but they do not provide instructions about the interpretation of the results.However, the ATS clinical practice guideline for the interpretation of F ENO defines high, intermediate, and low levels, with cutoffs of >50, 25-50, and <25 ppb, respectively (20).Consequently, we selected a cutoff of ≥25 ppb for our composite variable to include individuals even within the intermediate F ENO level to ensure that we did not exclude anyone with eosinophilic airway inflammation.This choice was driven by the fact that our study focused on elite athletes who engage in extreme physical performance, and even minor disturbances may have a significant impact on performance.Moreover, our findings indicate that F ENO levels are elevated in individuals with asthma, and the Youden's index based on the ROC analysis determined an optimal cutoff point for F ENO at 20 ppb for predicting the diagnosis of asthma.Although it is lower than that recommended in the ERS guidelines, this cutoff was calculated to optimize the sensitivity and specificity, whereas the guideline's higher cutoff was fitted to yield a high specificity to confidently rule in the diagnosis of asthma in symptomatic participants.Our F ENO results and the calculated cutoff are in line with a series of previous studies (46)(47)(48)(49).Nevertheless, designing prospective studies to assess other cutoff values of F ENO for diagnosing asthma in elite swimmers may be of interest.Allergy is the most common pathophysiological characteristic of asthma in the general population, almost 60% of asthma cases are attributable to atopy, according to a nationwide survey conducted in the United States involving more than 12,000 participants (50).In the pathogenesis of type 1 hypersensitivity, IgE plays a critical role in mediating the degranulation of mast cells and basophils and subsequently leads to the recruitment and activation of eosinophils (51).The level of total serum IgE correlates with asthma in adults and pediatric patients, as has been proven in large-scale international studies (52,53).We also found that IgE levels were elevated in asthmatic participants compared with nonasthmatic participants and that it could distinguish between those with and without asthma.However, our cutoff value calculated using Youden's index was lower than that reported previously (48,54).Furthermore, the skin prick test is also associated with allergic asthma, and we found that more than half of the asthmatic participants were sensitized to at least one inhaled allergen.This is in line with the results of the Third National Health and Nutrition Examination Survey in the United States, which found that 56.3% of patients with asthma also had atopy (50).
Eosinophil granulocytes are important participators of allergic inflammation, and their role in the pathogenesis of asthma, i.e., in airway remodeling, has been proven long ago (55).Identifying eosinophilic airway inflammation is possible through the analysis of induced sputum or bronchoalveolar lavage, but it is burdensome to both the patient and the healthcare system and is thus rarely feasible (56).However, BEC may reflect the airway compartment, and its measurement is widely accessible.Furthermore, according to large-scale epidemiological studies, elevated BEC is associated with poor disease outcomes and frequent exacerbations (57).Therefore, several tailored biological treatments of asthma aim to lower BEC and thus ameliorate eosinophilic inflammation through targeting key inflammatory cytokines, e.g., interleukin (IL)-5, IL-4, and IL-13 (29,58).However, we did not find differences in either the absolute eosinophil count or eosinophil ratios between asthmatic and nonasthmatic participants.A possible explanation may be that our study did not have a sufficient sample size to detect the difference because according to a study involving more than 130,000 asthmatic patients, only about one-fifth of the patients present with elevated BEC (57).Furthermore, swimming may also influence airway eosinophils, although existing research on this matter has shown mixed results.Bougault et al. (36,59) reported elevated eosinophil counts in induced sputum and bronchoscopic samples of nonasthmatic swimmers, whereas others found no significant differences between swimmers and nonswimmers (60,61).However, the evidence suggests that swimming may not significantly impact BEC levels (62,63).
The strength of our study is that we were able to enroll the majority of the National Swim Team of Hungary, which allowed us to perform the analyses with a sufficient sample size to draw certain conclusions about this highly specific population.Furthermore, our scoring system showed a fairly strong association with the diagnosis of asthma, and the required tests are widely accessible in secondary care and sometimes even in primary care.However, our study has limitations as well.First, because of the single-center setting and the single ethnicity of our population, our results may not be universal and need further validation in multicenter studies with the enrolment of mixed ethnicities.Furthermore, although the sample size is substantial considering the relatively small population of elite water sports athletes in our region, it may still be insufficient to comprehensively explore various associations beyond the primary focus of our study.Moreover, we are also aware that the ERS guideline recommends against the combined use of F ENO , BEC, and serum IgE in the diagnosis of asthma as the accuracy of the combined test may not be substantially higher than of the individual tests; however, their recommendation is conditional and is based on one single-center study that assessed a different population (14).Additionally, we involved in our combined system the skin prick test, which may increase the diagnostic accuracy.

FIGURE 3 -
FIGURE 3-ROC curves of BEC, blood eosinophil percentage, total IgE concentration, and F ENO to predict asthma diagnosis.

TABLE 1 .
Demographic and clinical parameters.
Data are presented as mean ± SD or median (interquartile range) and are compared with unpaired t-test or Mann-Whitney U-test.Statistically significant differences ( P < 0.05) are highlighted in bold.BMI, body mass index; DL CO , diffusing capacity for carbon monoxide; FEF, forced expiratory flow between 25% and 75% of FVC; FEV 1 , forced expiratory volume in the first second; ΔFEV 1 , change in FEV 1 during reversibility testing (% change representing absolute change); IU, international unit; KL CO , transfer coefficient of the lung for carbon monoxide; ppb, parts per billion; WBC, white blood cell count.BIOMARKER ASSESSMENT IN ASTHMATIC ATHLETESMedicine & Science in Sports & Exercise ® 1259

TABLE 2 .
Cross-tabulation of the proposed scoring system and positive asthma diagnosis.

TABLE 3 .
Results of binary logistic regression model.P values <0.05 are highlighted in bold.The composite biomarker variable is produced from sum of the dichotomized values of BEC, total IgE concentration, F ENO , and positive prick test.FEV 1 , forced expiratory volume in the first second of exhalation.