Exercise-induced asthma (EIA) is the term used to define the transient narrowing of the airways[1] that follows vigorous exercise[2,3] in a dry environment.[4,5] EIA most commonly occurs in those with clinically recognized asthma with a point prevalence of around 50–65% of those being treated for asthma.[6,7] Exercise-induced bronchoconstriction (EIB) is the term proposed to describe the same airway narrowing in people who do not have other signs and symptoms of asthma.[8] EIB has become increasingly recognized in school children,[9,10] college athletes[11] elite athletes,[12–14] and defense force recruits.[15,16] The common finding in these groups is that these individuals usually have normal spirometry at baseline.

The stimulus for the airways to narrow is the evaporative loss of water in conditioning the inspired air. The mechanism whereby the loss of water provokes airway narrowing is thought to involve the thermal[17] and osmotic consequences of dehydration.[18] Because EIA can occur when hot dry air is inspired,[19–21] the osmotic consequences are thought to be more important than the thermal ones under the inspired air conditions of temperate climates.[22,23]

The findings of epidemiologic studies suggest that EIB is likely to be one of the first measurable signs of asthma, preceding other features such as reduction in lung function and symptoms and airway hyperresponsiveness to methacholine (see figure 1).[9,10,24–27] Airway hyperresponsiveness to methacholine was predicted in adolescence, in children with EIB but without clinical asthma, 4.6 years earlier.[27] In another study, 32 of 55 children with EIB had developed asthma within 6 years compared with only 11 of 55 who did not have EIB at the time of study.[26] There was also a significantly higher prevalence of atopy and respiratory-related illness in those who had EIB.[26] Longitudinal studies of the same geographical location show that the prevalence of EIB is increasing.[28,29] EIB is more common in an urban environment than a rural one.[30,31]

Fig. 1
figure 1

Schema for the progression of indicators of asthma during childhood. The schema was based on studies in children who had associated early atopy with the later development of asthma,[24] and associated severity of atopy with severity of exercise-induced asthma (EIA).[25] Children with early EIA were likely to have asthma 6 years later,[26] and EIA at 7 years of age predicted airway hyperresponsiveness (AHR) to methacholine at 11.6 years.[27] Furthermore, EIA was noted in school children before symptoms and changes in lung function associated with asthma had developed.[10]

Full size image

Moderate to severe EIA is accompanied by arterial hypoxemia,[32] with some gas trapping and hyperinflation.[33] These abnormalities do not occur if EIA is prevented by a pharmacologic agent.[32,33] For this reason, safe and effective prophylactic therapy is recommended for the prevention of EIA. The findings of studies investigating the effect of long-term treatment with inhaled corticosteroids (ICs) show that EIA is the last easily measurable sign of asthma to disappear after treatment.[34] Resolution of EIA lags well behind baseline FEV1, symptoms, and peak flow variability, which are all improved at lower doses of ICs relative to those required to control EIA (figure 2).[34–36] It seems that EIA is one of the first indicators of asthma (figure 1) and one of the last to resolve with treatment (figure 2). For this reason, single-dose treatment of EIA must be considered in conjunction with long-term treatment for the underlying disorder of EIA (i.e. inflammation of the airways).[37–39]

Fig. 2
figure 2

Schema to illustrate response of various outcome parameters of asthma severity with increasing doses of inhaled budesonide[34,35] (reproduced from Barnes et al.,[36] with permission). Exercise (FEV 1 ) = the fall in FEV1 on exercise; FEF 25–75 = forced expiratory flow between 25% and 75% of forced vital capacity; NO = nitric oxide concentration in expired air; symp. = both morning and evening asthma symptoms.

Full size image

Over the last 40 years, many publications have reported the safety and efficacy of different drugs for EIA, particularly β2-adrenoceptor agonists (β2-agonists), cromolyn sodium (sodium cromoglicate), and nedocromil.[40–44] During this time period, the propellants have changed in the pressurized metered dose inhalers (PMDIs) and dry powders and delivery devices have become more commonly available. The β2-agonists have become more receptor specific and act for a longer duration and a new class of drugs the leukotriene receptor antagonists (LTRAs) have been introduced. In recent years, newer drugs and combination therapy with improved adverse effect profiles have diverted our attention away from proven therapies. PharmacogenomICs, a new field of study, has shown that the same drugs may not lead to the same outcomes in all people with the same disease;[45] the recent publications on β-receptor polymorphisms are a good example of this.[46,47] This new knowledge means that the simplistic approach to preventing EIA that we used in the past must now be tempered by an awareness of the relative advantages and disadvantages of the different drugs available for the prevention of EIA.

1. Development of the Single-Dose Approach to Exercise-Induced Asthma (EIA)

The first reports of the efficacy of a single dose of medication to prevent EIA were in children in the early 1960s.[48] The inhaled medications used were isoproterenol (isoprenaline), a relatively non-selective β2-agonist, and adrenaline (epinephrine). The preventative effects of single doses of drugs on EIA was later developed and refined to demonstrate duration of action in children by Godfrey, Connelly-Wilson, Silverman, and Konig[49–52] at the Brompton Hospital in London in the early 1970s. Systematic studies using a randomized crossover design with placebo controls were carried out for a number of agents including cromolyn sodium, albuterol (salbutamol), theophylline, and the anticholinergic agent atropine.[49–51]

1.1 Cromolyn Sodium (Sodium Cromoglicate)

At that time these pediatricians were particularly interested in the benefits of the new drug cromolyn sodium in EIA. The cromolyn sodium 20mg dose was combined with a low dose of isoproterenol 100μg in the earliest studies in EIA.[52] Combination therapy was unusual at the time but it proved to be very effective in preventing EIA in children.[40] A single dose of cromolyn sodium (20mg delivered by a Spinhaler™)Footnote 1 inhibited EIA in most of the 80 children reported by Anderson et al.[53] but the protection lasted <2 hours.

Placebo-controlled crossover studies of the effects of a single dose of β2-agonist, cromolyn sodium, the anticholinergic agent ipratropium bromide, and theophylline were also carried out in adults with exercise-induced asthma.[54–57] The findings in adults were similar to children: β2-agonists and cromolyn sodium, given as aerosols, were effective in preventing EIA, but only for a short duration (usually <2 hours). The short-acting β2-agonists (SABAs) albuterol, terbutaline and hexoprenaline taken as tablets were effective bronchodilators but they did not prevent the post-exercise fall in lung function in 80% of the individuals.[41,54,55] Similarly, the anticholinergic agent ipratropium bromide induced bronchodilatation but prevented EIA in only around 20% of the individuals.[41] In contrast, cromolyn sodium did not induce bronchodilatation but it did provide a significant protective effect on the fall in lung function after exercise.[41,53]

From these studies it became clear that: (i) relaxing bronchial smooth muscle was not an essential requirement for a drug to prevent EIA; (ii) most drugs had a protective effect for <2 hours; (iii) drugs given as aerosols were more effective than tablets; (iv) drugs that were purported to prevent the release of inflammatory mediators from mast cells such as cromolyn sodium and the β2-agonists were very effective in inhibiting or preventing EIA; and (v) drugs that had a dual action (i.e. preventing mediator release and relaxing bronchial smooth muscle), such as the β2-agonists, were the most effective in inhibiting EIA.

1.2 Inhaled Corticosteroids

ICs were introduced for the treatment of asthma during the 1980s; the effect of single doses of agents on EIA was often investigated with concomitant use of ICs. The results of these early studies were important to our understanding of the pathogenesis of EIA and the role of β2-agonists, yet the findings were not really appreciated at the time. First, it was demonstrated that daily treatment with ICs alone could reduce the severity of EIA substantially, so it became evident that inflammation of the airways was an important determinant of the presence and severity of EIA.[58,59] Secondly, only 37μg of a β2-agonist, i.e. <10% of the standard dose used today, was required to prevent EIA with concomitant use of ICs.[58] This finding must make us consider why the recommended dose of 500μg of terbutaline used to prevent EIA is so high today.[60] For most of the 1980s the recommendations for the prevention of EIA included the prophylactic use of a standard dose of SABA or cromolyn sodium given by inhalation 15–30 minutes before exercise.

The concept of combining drugs to prolong the duration of protection against EIA was investigated in formal trials involving the use of a β2-agonist in combination with cromolyn sodium[61] and with ipratropium bromide.[62] During the 1990s, ICs had become more common in the treatment of asthma, and many patients no longer had EIA and did not need prophylactic treatment before exercise.[7] This important fact was not interpreted in the way that it would be with today’s knowledge (i.e. no EIA is a good index of benefit from ICs). Rather, at the time, exercise was interpreted as being insensitive for the identification of airway hyperresponsiveness (AHR) in patients known to have asthma.

1.3 Long-Acting β2-Adrenoceptor Agonists (LABAs)

Long-acting β2-agonists (LABAs) were introduced for the treatment of asthma in 1991. Some of the first publications reported the effect of a single dose of salmeterol for the prevention of EIA in adults and children.[63–66] The first two articles[63,64] on the subject revealed information, the significance of which is only now coming to be recognized by the wider medical community. In the study by Anderson et al.,[63] a single dose of salmeterol 50μg was administered for the first time in exercise-induced asthma; the duration of the protective effect was substantially less than suggested by the dosing regimen (12 hours). At 6.5 hours after administration of the drug, only 3 of 17 individuals had complete protection (<10% fall in FEV1) and 11 of 17 had clinical protection (<50% fall in FEV1 with the active drug compared with the placebo). The reason for this unexpected finding was suggested by tolerance to the protective effects of the drug as reported by Ramage et al.[67] On the first day of their study, after the first single dose of salmeterol, the protective effect of the drug against EIA was noted for 6 and 12 hours, i.e. protective for the full dosing period. However, when salmeterol was taken daily over a period of 1 month, the duration of the protective effect was reduced to 6 hours, a time also observed for many individuals taking β2-agonists daily in the initial study.[63] This tolerance is now well recognized and can develop with the daily use of β2-agonists, either short or long acting, alone or in combination with ICs.[68,69] Tolerance does not appear to be a problem when these drugs are used in single doses three times per week to prevent EIA.[70]

The reasons for the development of ‘tolerance’ to the protective effect are discussed elsewhere in several articles.[60,71,72] Another review in children[43] also highlights the inconsistent time of protection afforded by the β2-agonists in individuals with EIA within the dosing interval. In brief, these reviews concluded that the daily use of β2-gonists, either short or long acting, led to a reduction in duration of the protective effect on EIA,[67–69] a prolongation in the time of recovery from bronchoconstriction,[73–75] and an increase in sensitivity of the smooth muscle to contractile agents.[76,77] The likely reason that these effects are not reported more often is that the protocols have required time off treatment prior to study entry.

During the late 1990s, the LABA salmeterol was increasingly used in the treatment of asthma and the term ‘symptom controller’ was coined. A second LABA, formoterol, also became available and, like salmeterol, was shown to be effective in preventing EIA when given in single doses to either children[78,79] or adults.[80] Some investigators performed dose-response studies with these LABAs in an attempt to optimize the dose required to prevent EIA.[81–84] The different dry powder devices with variable resistance on inspiration were also investigated and shown to be effective even in young children with asthma.[6,85,86]

1.4 Nedocromil

During this period there was a decline in the use of drugs such as cromolyn sodium to prevent EIA. The reason for this probably relates to the much lower dose used when the PMDI was introduced, particularly in the US. The dose available per inhalation was only 1mg; thus, the standard two inhalations delivered 2mg, i.e. only one-tenth of the dose used in the capsule and delivered by the Spinhaler™. This low dose did not appear to provide much protection in patients with moderate to severe EIA.[61] Although the potential for superior delivery and deposition of the drug into the lower airways using a PMDI compared with the Spinhaler™ is acknowledged, the lower dose would have increased the failure rate of cromolyn sodium, which, in adults, appears to be very dose dependent (figure 3).[87] In many countries PMDIs were available at 5mg per inhalation, so that increasing the dose to the original 20mg was not difficult and administration of cromolyn sodium by PMDI was recommended as prophylaxis for EIA.[44,57,87] However, some investigators have found low doses of cromolyn sodium to be effective in children with EIB.[88]

Fig. 3
figure 3

The mean fall in FEV1, expressed as a percentage of the pre-exercise values, in 11 adult patients with asthma exercised on a treadmill for 6–8 minutes and inspiring air in temperate conditions (20–22°C, relative humidity 40–60%). Individuals exercised for 30 minutes after inhaling placebo and cromolyn sodium (sodium cromoglicate) 2, 10, and 20mg. Four, nine, and ten of the 11 individuals had a fall in FEV1 that was <47% of that observed with placebo with the cromolyn sodium 2, 10, and 20mg doses, respectively (drawn from data).[87] ? indicates the response that would be likely if the dose was 40mg.

Full size image

Another reason why cromolyn sodium fell into decline was associated with the launch of nedocromil, a drug with a similar efficacy profile but more potent than cromolyn sodium. In single 4mg doses, nedocromil was demonstrated to be effective against EIA in both children[89] and adults.[90–92] As with cromolyn sodium and the SABAs such as salbutamol and terbutaline, the duration of the protection afforded by nedocromil was short (usually <2 hours).[93]

The studies on nedocromil given in a single dose (range 1–8mg) 15–60 minutes before exercise have been subject to a meta-analysis using the Cochrane system.[92] The rate of protection was 51% and the time to recover normal lung function was <10 minutes after nedocromil compared with >30 minutes after placebo. The protective effect was <2 hours in most cases, but it is important that the most benefit occurred in those individuals with the most severe EIA.[92]

In another analysis comparing the benefits of cromolyn sodium 10mg and nedocromil 4mg in EIA, no significant differences were found in relation to the maximum fall in FEV1 and the protection afforded by the drugs in children and adults.[94] The data for one study in children are illustrated in figure 4.[89] The appealing features of cromolyn sodium and nedocromil sodium are that these drugs are active immediately and effective at the first dose, they are well tolerated, and they cause no adverse effects of significance. Doses can be titrated to suit the individual (figure 3).[57,87] Their mode of action, from experiments performed in vitro, appears to involve the prevention of release of inflammatory mediators from mast cells and prostaglandin D2 (PGD2) in particular.[95] It has also been suggested that they prevent the release of neuropeptides from sensory nerves.[96] Further they are effective against the most common provocative stimuli, for example allergens, cold air, exercise, non-isotonic aerosols, fog, sulfur dioxide, and metabisulfite in food and wine. The disadvantages are the short duration of action and the need to identify the effective dose.

Fig. 4
figure 4

Individual values for the percentage fall in FEV1 in 12 children with asthma after a baseline exercise test and after treatment with nedocromil (NED) 4mg, cromolyn sodium (sodium cromoglicate [CRO]) 10mg, and placebo (PL) administered with a metered-dose inhaler without or with the spacer (Fisonair™). Exercise was performed 30 minutes after the aerosol had been taken (drawn from data).[89]

Full size image

1.5 Leukotriene Synthesis Inhibitors

Over the years it became clear that mast cell mediators were important to understanding the etiology of EIA. These mediators include histamine, leukotrienes, and prostaglandins, specifically PGD2. Unlike histamine, which is a preformed mediator, the leukotrienes and prostaglandins are synthesized de novo after activation of the mast cell. Investigators reasoned that preventing synthesis of these mediators (5-lipoxygenase inhibitors, cyclo-oxygenase inhibitors) or blocking the response at the receptor level could reduce the severity of EIA. By the early 1990s new drugs that could inhibit production or block response to these mediators became available for the treatment of asthma with the potential to prevent EIA. These included inhibitors of the enzyme 5-lipoxygenase[97,98] and the LTRAs.[99,100] The major advance in treatment was that these agents could be given as a tablet. Until that time, aerosols had been the mainstay of single-dose treatment for the prevention of EIA for 30 years.

Although the 5-lipoxygenase inhibitors (zileuton)[101,102] at single doses have a significant inhibitory effect on EIA, they have not become widely prescribed for this use. In contrast, the LTRAs are recommended for EIA, particularly in children.[103] This is likely to continue as there are now results from studies confirming that single doses of LTRAs are as effective as regular dosing in the prevention of EIA.[102,104] The acceptance of the LTRAs as beneficial for use in EIA came about because the traditional index for measuring severity of EIA (percentage fall in FEV1) was complemented by an analysis of the FEV1 time curve for 60 minutes following exercise (figure 5).[99,105] This analysis showed that although LTRAs do not completely prevent the post-exercise fall in FEV1 (protection was about 60%), they do significantly reduce the time to recover to baseline FEV1. In the study by Melo et al.,[105] the maximum percentage fall in FEV1 was reduced from 35.1% ± 2.6% to 17.3% ± 2.4%, the area under the curve was reduced from 868.0 ± 103.8 FEV1 % change/min to 267.8 ± 42.7 FEV1 % change/min, and the time of recovery of FEV1 to within 10% of the pre-exercise value was reduced from 30.9 ± 4.0 minutes to 6.9 ± 1.1 minutes (p < 0.05). As the major reduction in FEV1 occurs after cessation of exercise, rather than during exercise, the fall is not considered to limit exercise performance. The fast recovery in lung function also enables the person to exercise again within a short period of time (10 minutes).

Fig. 5
figure 5

The mean fall in FEV1 after 6–8 minutes of cycling exercise, expressed as a percentage of the baseline FEV1 in 22 children aged 7–16 years treated with montelukast 5 or 10 mg/day or placebo for 1 week. The values were significantly lower after montelukast treatment compared with placebo immediately after exercise and at 5, 15, 30, and 60 minutes (reproduced from Melo et al.,[105] with permission). * p < 0.05 vs placebo.

Full size image

At the time of the major studies with cromolyn sodium and nedocromil, it was not usual to analyze the area under the FEV1 time curve; rather, the time to recovery to baseline was often commented upon. However, if analyzed in a fashion similar to that used with the LTRAs, the data so often illustrated in publications suggest that cromolyn sodium and nedocromil would provide about 60% protection,[87] similar to that with montelukast. Importantly, both classes of drug enhanced recovery time to pre-exercise lung function.

1.6 Cyclo-oxygenase Inhibitors

The drugs that inhibited cyclo-oxygenase and thus the production of prostaglandins, such as flurbiprofen and indomethacin (as an aerosol), were also shown to reduce severity and enhance recovery, suggesting that prostaglandins also contribute to sustaining the airway response to exercise.[106,107] The thromboxane B2 receptor antagonists were not shown to be effective at all, although they did inhibit responses to inhaled PGD2, the major prostaglandin implicated in EIA.[108,109] These agents are not used in either the treatment of asthma or the prevention of EIA, but they have given insight into disease pathogenesis. For example, inhibitors of the enzyme cyclo-oxygenase, also referred to as NSAIDs, prevent refractoriness developing in the post-exercise period. This period of refractoriness to exercise occurs in about 50% of patients with asthma and appears to be dependent on the release of ‘protective’ prostaglandins, possibly prostaglandin PGE2.[110–112] Inhalation of dinoprostone, a PGE2, attenuates EIA.[113] It should be noted that this protective effect is not by functional antagonism of the contractile effects on bronchial smooth muscle, as the responses to methacholine are unaffected by PGE2.[113] PGE2 is also released in response to leukotriene (LT) D4, another mediator implicated in EIA, and this may help to explain the development of a refractory period after exercise.[114] In summary, the cyclo-oxygenase inhibitors probably do more harm than good in terms of EIA.

1.7 Histamine Receptor Antagonists

Inconsistencies in the findings of the effects of histamine antagonists in EIA have been discussed recently.[115] Patients with more severe EIA or those in whom a more intense ventilatory stimulus has been used appear to benefit the most from these agents, whereas those with milder responses do not.[116–121] The combination of a histamine and an LTRA has also been investigated in EIA, and there does not appear to be any significant additive effect. Children with very mild EIA (maximum percentage fall in FEV1 of 15.3% ± 2.9% 2 hours after placebo and 18.7% ± 2.8% 12 hours after placebo) were given montelukast 5mg together with loratadine 10mg.[121] The AUC0–30 min (FEV1 % change/min) and the maximum percentage fall in FEV1 were not significantly different at 2 hours. At 12 hours, the maximum percentage fall in FEV1 was 43.6 ± 9.34 FEV1 % change/min after placebo versus 15.03 ± 5.4 FEV1 % change/min after montelukast and 7.76 ± 6.14 FEV1 % change/min after the combination (p < 0.001). The maximum percentage fall in FEV1 at 12 hours with the combination of montelukast and loratadine was 9.5% ± 2.55% (p < 0.02). It seems that in these children, 2 hours was not sufficient time to achieve the required blood level of the drug to have a significant effect; a second dose may have achieved the desired effect.[99,105] The combination of zafirlukast and loratadine was also studied in adults, but not in single doses.[117] No additional protection was provided by loratadine 10mg twice daily in combination with zafirlukast 80mg twice daily for the group analysis. Five of 16 individuals, however, did appear to benefit from the combination therapy when it was taken twice daily for a week.

Some investigators have failed to implicate the release of inflammatory mediators from mast cells in the airways by means of bronchoalveolar lavage after exercise,[122] although others have been more successful after eucapnic hyperpnea of dry air.[123] However, there is so much indirect evidence to involve mast cell mediators in EIA from studies of drugs that the quest to overcome the technical problems continues. Exercise has been shown to be associated with significant increases in urinary excretion of the mediators LTE4 and the metabolite of PGD2, 9α,11β-PGF2,[124,125] and 50% of asthmatic patients with EIA have a significant increase in arterial plasma histamine levels measured at the time of maximal airway narrowing from exercise.[126,127] Mannitol, an osmotic stimulus that has been used as a surrogate for exercise, has also been shown to have a similar profile to exercise, with respect to both the preventative action of single doses of drugs and the release of mediators.[128–130] As with exercise, the levels of LTE4 and 9α,11β-PGF2 are increased in urine after mannitol inhalation.[131] It has recently been reported that the increase in urinary excretion of 9α,11β-PGF2 (but not LTE4) is prevented by pre-medication with cromolyn sodium 40mg or the LABA formoterol 24μg.[130] In this study, the mannitol was administered in a sufficient dose to induce a reduction in FEV1 similar to that observed with exercise (31.8% ± 10.2%). This is the first in vivo evidence that these drugs are able to reduce mediator release from mast cells and confirms the earlier in vitro findings.[95] The inhibition of the release of the PGD2 metabolite was associated with inhibition of the airway response to mannitol. From these studies, it would now be of interest to know the combined effect of single doses of cromolyn sodium and montelukast in EIA.

2. EIA in the 21st Century

Over the last 5 years, two new combination treatments have become widely prescribed for treatment of asthma. Both involve a LABA and an ICs. One is the combination budesonide/formoterol at respective doses of 100µg/6µg, 200µg/6µg, and 400µg/12µg (Symbicort®). The other is a combination of fluticasone propionate at a dose of 50, 100, 250 or 500µg with salmeterol (50µg) [Seretide®]. These drugs are given twice daily. Both these classes of drug have a beneficial effect on EIA, and one might predict a declining need for single-dose medications to prevent EIA. However, as discussed in section 1.3, tolerance to the protective effects of LABA develop with daily use, even in those taking ICs. This means that there is likely to be a significant number of people requiring additional single-dose prophylactic therapy for EIA or rescue therapy for breakthrough EIA between dosing schedules. It would be useful to know the percentage of adults and children taking a LABA daily who are finding themselves in need of additional bronchodilator or prophylactic therapy. In precautions on the use of LABAs listed in the Australian MIMS,[132] it is acknowledged that there may be acute symptoms in patients taking LABAs and the use of SABAs is recommended as rescue therapy.

It is now recognized that 16% of the Caucasian population and 25% of the African American population in the US have a polymorphism of the β2 receptor, which results in poorer outcomes of treatment with LABA alone or in combination with an ICs.[46,133] The poorer outcomes may explain why there has been a small but unexpected increase in the number of deaths from asthma[134] and the number of exacerbations requiring hospitalization[135] in people taking LABA relative to placebo.

There are other aspects to the daily use of β2-agonists that may change usage and prescribing habits. Individuals seeking approval to enter into some occupations or sports such as recreational diving or international sporting competitions need to ensure that they inform the relevant officials and organizers that they are using β2-agonists. For example, the New South Wales police force in Australia will permit recruits with asthma after demonstrating that they can exercise without prophylactic treatment with a β2-agonist to control EIA. The potential recruit is permitted to take ICs to control airway inflammation and reduce the likelihood of EIA.[136]

There was concern over the marked increase in reports of β2-agonist aerosol use before an event from 3.6% at the Olympic Games in Atlanta, Georgia, USA in 1996 to 5.5% in Sydney, Australia in 2000. There was also concern that many athletes using β2-adrenoceptor agonists were underusing ICs to treat their asthma. These factors prompted the International Olympic Committee (IOC) Medical Commission to request objective evidence for asthma or EIB (the only medical indications for the inhaled formulation of β2-agonists) for the Winter Games in Salt Lake City, Utah, USA in 2002.[137] Although the number of athletes requesting permission to use β2-agonists was expected to increase for the Salt Lake City Games relative to the Nagano Games in Japan in 1998, this was not the case. The reason for this is not clear but may relate to the team doctors and athletes themselves becoming more cautious about prescribing or using a β2-agonist without documented EIA or to the relatively short notice given. The higher prevalence of β2-agonist use at the Nagano Games relative to the Salt Lake City Games by the same teams (Sweden 16.5% vs 8.3%, USA 16.9% vs 12.9%, Canada 10.3% vs 4.5% of athletes) suggested that either EIA had been over-diagnosed in elite athletes and/or that asthma had been under-treated with ICs. Some athletes may simply have decided that using a β2-agonist before exercise did not alter their respiratory symptoms. The requirement to submit evidence for the need for a β2-agonist by inhalation was also adopted by the International Amateur Athletic Federation for the World Championships in Paris in 2003, and this body continues to require this evidence. The same requirements were also in place for the Summer Olympic Games in Athens, Greece in 2004. If the IOC policy for the Olympic Games becomes a national, state, or local community policy for sporting competition, then the medical community needs to re-assess the evidence on which they base their prescribing habits for both asthma and EIA. The characteristICs of a patient’s asthma need to be considered together with the requirements of sporting bodies when prescribing treatment for EIA. The MIMS manual for pharmaceutical agents provides a useful index for the physician in terms of drugs and sport.[132]

3. Recognizing the Problem

Clinically, EIA is best recognized by a positive history of breathlessness 5–10 minutes after exercise. The breathlessness at this time should be worse than that experienced during exercise. This history distinguishes a person who is breathless as a result of having reached their maximal expiratory airflow (elite athlete or patient with chronic airflow limitation) from one whose airways narrow acutely after exercise has ceased. Most people who exercise strenuously feel very breathless in the last moments of exercise but recover within 10 minutes. Fifty percent of elite athletes who complain of breathlessness and chest tightness during exercise do not have EIB after exercise.[12] The same findings have been made in school and college athletes.[10,11] The best tool for confirming the diagnosis of EIA is to measure the response to exercise or a surrogate of exercise.[6,138,139]

3.1 Abnormal Spirometry

It is also important to know if a person complaining of EIA has normal spirometry at rest. If they do not, then it is useful to find out if the patient has an acute improvement in response to a bronchodilator (i.e. an increase in FEV1 >15% of baseline or 12% of predicted FEV1). It is also useful to seek more information about underlying asthma (waking breathless at night, allergies to cats and dogs). If they do have a significant response to a bronchodilator, treatment of the underlying inflammation with a course of ICs is recommended. Attention should be given to identifying and removing, where possible, obvious sources of allergens, particularly in the bedroom. Allergen exposure outdoors is thought to be an important factor in the development EIA in elite athletes, who may spend many hours training outdoors.[140] Similarly, there are indoor irritants such as chlorine,[141] and ultrafine particles can predispose people to airway hyperresponsiveness.[142]

When trying to establish a history of severity of asthma, it is important to take into account the number of inhalations of β2-agonist a person may already be taking before exercise. Some guidelines have ignored or specifically excluded the extra doses of β2-agonists that are taken before exercise when assessing asthma severity. In doing so, there is the potential for the severity of asthma to be underestimated in some patients.

3.2 Normal Spirometry

If a person has a history of EIA and no other symptoms (including vocal chord dysfunction) and normal spirometry, then there is an opportunity to use single-dose treatment as required before exercise. Nowadays it is most common for a person with this history to be prescribed a β2-agonist before exercise. This approach, though attractive to many, is palliative only. It does not address the question as to why the patient has EIA (airway inflammation) and it does not acknowledge that EIA may be the first measurable sign of developing asthma (see figure 1). It would be better to know the severity of EIA before embarking on prophylactic treatment with a β2-agonist. Many people may have good lung function but a moderate or even severe airway response to exercise. Inexpensive meters ($US50) that measure FEV1 are now available. Thus it should not be considered impossible or even difficult to document the conditions under which a person shows the symptoms of EIA. The measurements of FEV1 are made before and three or four times after 6–8 minutes of exercise of sufficient intensity to raise the heart rate to 90% of predicted maximum for the last 4–6 minutes. In the field, a positive response to exercise is a fall in FEV1 of ≥10% or 15% of baseline. When tests are normal and the individual is still having difficulty with exercise, vocal chord dysfunction should be considered and referral for confirmation of this is advised.

4. Recommendations for Treatment

4.1 Inhaled Corticosteroids

It needs to be kept in mind that regular treatment with ICs alone is very effective in reducing the severity and number of episodes of EIA in children[34,143,144] and adults.[145] Many pediatricians are hesitant about using ICs for long periods to control asthma let alone to control EIA because of potential adverse effects.[146] Indeed, treatment and prevention of EIA has become an issue in children;[43,44,60,103] perhaps some compromise is needed with initial treatment with ICs and withdrawal after 8–12 weeks.

The recommendation for individuals considered to have moderate to severe EIA is to use a flow or volume meter to measure the response to the exercise task that is causing the symptoms. If the value falls by ≥15% of their baseline after exercise, they should be advised to take ICs twice daily and repeat the test after 1 week. If the repeat test is successful, in that the severity of EIA is reduced, the patient may be more inclined to complete a course of ICs treatment, say for 8 weeks. If there has been no improvement after 1 week, the dose of ICs could be increased, after checking inhaler technique and adherence to dose, and the test repeated after 4–8 weeks.

The advantages in taking the ‘treat the underlying asthma with ICs’ approach to EIA is that the majority of subjects will no longer require prophylactic treatment for EIA after 8–12 weeks. In a study of children with normal values for FEV1 (mean 101% ± 10%), Jonasson et al.[144] showed that the maximum percentage fall in FEV1 after exercise was reduced from 25.3% ± 10.8% to 9% ± 7% after 12 weeks of taking budesonide 200µg once daily. Similar results were reported by Hofstra et al.[143] in children taking fluticasone propionate 100 and 200µg. It is not known for how long EIA would be suppressed after withdrawal of corticosteroids, but it could be months judging from the data of a study using mannitol as the provoking stimulus.[147] Those taking ICs alone should be advised to carry rescue bronchodilator therapy with them, and this would be expected to work quickly and well in a person not taking a β2-agonist on a daily basis.

The disadvantage to this approach is that ICs have the potential to cause unwanted adverse effects, something that many patients with asthma and their parents may want to avoid risking. Further, many general practitioners and pediatricians do not wish to prescribe ICs for the treatment of EIA even in low doses.

The myth that EIA is unaffected by acute pre-medication with ICs has been countered by results of one study [148] that used a single 1mg dose of fluticasone propionate and demonstrated a significant reduction in EIA 4 hours after treatment. However, the use of single large doses of ICs to prevent EIA is not recommended.

4.2 Cromolyn Sodium and Nedocromil

If patients with normal spirometry would prefer to take a drug other than ICs to prevent their EIA, then the recommendation should be inhalation of either cromolyn sodium 2, 10, 20 or 40mg or nedocromil 4 or 8mg immediately before exercise (figure 3 and figure 4).[57,87,89,91,94,149]

The advantage in using these drugs is that they work effectively and immediately from the first dose and do not require any preplanning: the patient can inhale and ‘run’. These drugs can be used safely and frequently without the development of tolerance to their protective effects over the long term. There have been no significant adverse effects reported for either of these drugs. The doses can be adjusted, as some individuals require 20 or 40mg of cromolyn sodium to prevent EIA whereas others may need only 2 or 10mg (figure 3).[57,87] These drugs also work to prevent allergen-induced asthma during exercise, cold air- or fog-induced asthma, and sulfur dioxide-induced asthma. The protection on the percentage fall in FEV1 is 50–60%, and the time to recovery is <10 minutes.[94] In addition, sporting bodies do not require notification of the use of either cromolyn sodium or nedocromil sodium before a sporting event.

The disadvantage with these drugs is that for some individuals they are ineffective and/or short acting. These individuals are best advised to carry rescue medication in the form of a β2-agonist. The dispenser needs to be cleaned regularly if the PMDI with the new chlorofluorocarbon-free propellants are used. The other major disadvantage for the use of cromolyn sodium is that some adults are likely to require multiple inhalations if the dose is only 1mg per inhalation; however, this dose is successful in children.[88] In this case if a dry powder formulation of 20mg is available in a capsule, this could be used instead of a PMDI. Although the PMDI formulation is no longer available in some countries, the 20mg capsule is available in Europe, North America, Australia, New Zealand, and Japan, and this availability is in keeping with the return in popularity of dry powder formulations.

It is a myth that cromolyn sodium and nedocromil need to be taken regularly or need to be taken 15–30 minutes before exercise. They are effective in a single dose immediately and on first administration,[50] and they do not need to be taken every day to be effective in EIA. Indeed, their preventative action usually lasts <2 hours, so they should be taken immediately before exercise.

4.3 Leukotriene Receptor Antagonists

If a person with normal spirometry would prefer to take a tablet or a parent would prefer a child with asthma to be premedicated before school, then a LTRA such as montelukast (10mg adults, 4mg children) or zafirlukast (20 and 80mg doses reported) may be recommended. These drugs have been shown to give significant protection from EIA, with montelukast having the longest duration of protection. The advantage of LTRAs is the long duration of protection after a single dose (figure 6)[102] and the ease of administration of tablets or syrup to children. Even if some EIA does occur, rapid recovery within 10 minutes can be expected.[105] There are no reports of tolerance with repeated administration of LTRAs over 8 weeks. There have been no reports of significant adverse effects. These drugs are also likely to reduce allergic reactions, and there are reports of montelukast being beneficial in those individuals with allergic rhinitis.[150] In addition, sporting bodies do not require notification of the use of leukotriene antagonists before a sporting event.

Fig. 6
figure 6

The mean fall in FEV1 after exercise, expressed as a percentage of the baseline FEV1 before exercise in ten patients (mean age 29 ± 3 years) with asthma treated with leukotriene receptor antagonists or placebo. Patients performed 4 minutes of exhausting work on a cycle ergometer while breathing frigid air. The first challenge took place 1 hour after ingesting montelukast 10mg, zafirlukast 20mg, or placebo and was repeated 4, 8 and 12 hours later (reproduced from Coreno et al.,[102] with permission).

Full size image

The disadvantage of LTRAs is that EIA is unlikely to be completely prevented, although symptoms are diminished. Overall, LTRAs provide the same degree of protection as the standard dose of cromolyn sodium or nedocromil. The time of onset of action is variable and a loading dose may be required. Cost may be an issue in a person exercising vigorously every day. Rescue medication for breakthrough EIA should still be available in the form of inhaled β2-agonists.

It is a myth that LTRAs need to be taken regularly to be effective in EIA, as they can be effective in a single dose.[102,104] Another myth is that leukotrienes do not play an important role in asthma: the leukotrienes act to sustain airway narrowing in response to exercise as well as a wide range of stimuli encountered during exercise including cold air, allergens, and sulfur dioxide.

4.4 Short-Acting β2-Adrenoceptor Agonists

If a person with normal spirometry would prefer to use a SABA before exercise, this could be recommended, but the patient should be informed that taking multiple inhalations of SABA on a daily basis is not advisable.

The advantage of using a SABA is the immediate onset of action (figure 7, figure 8, and figure 9), as a bronchodilator effect is not required in a person with normal spirometry.[151,152] In addition, the cost for most SABAs is low. Finally, SABAs offer >80% protection for most patients not taking β2-agonists every day.

Fig. 7
figure 7

The mean values for FEV1, expressed as a percentage of the predicted normal value, measured before and after each treatment and for 30 minutes after exercise in 27 young adults with asthma. Each individual performed 8 minutes of exercise on a cycle ergometer on four occasions while breathing compressed dry air. The first exercise test served as a control. Exercise was then performed in a randomized order following administration of placebo, albuterol (salbutamol) 200μg delivered from a pressurized metered dose inhaler (PMDI), and a dry powder preparation of albuterol 200μg from a Diskus™ inhaler (also known as the Accuhaler™) [reproduced from Anderson et al.,[6] with permission].

Full size image
Fig. 8
figure 8

Individual values for the percentage fall in FEV1 in 27 young adults who exercised for 8 minutes on a cycle ergometer while breathing dry air. Exercise was performed 30 minutes after taking placebo or albuterol (salbutamol; 200μg) from a Diskus™ inhaler (drawn from data).[6]

Full size image
Fig. 9
figure 9

The mean change in FEV1 after exercise, expressed as a percentage of the pre-exercise value in 25 adults (mean age 33 years) with exercise-induced bronchoconstriction. Exercise was performed 13 times and each study day was separated by 48 hours. Individuals performed the exercise on a cycle ergometer for 6 minutes while breathing cold dry air. The intervals between inhaling the medication and starting exercise were 5, 30, and 60 minutes. There was no difference in the protective effect afforded by salmeterol (50μg), formoterol (12μg) or terbutaline (500μg) at any time point, and the drugs had a rapid onset of protection (drawn from data).[151]

Full size image

The disadvantage of using a SABA is that the protective effect lasts only 1–2 hours, requiring further treatment if exercise is to be performed many times in a day. Frequent administration of a SABA can lead to enhanced severity of EIA.[153,154] If severe EIA occurs, extra doses of rescue medication will be required for those who use β2-agonists daily.[73] Many sporting bodies now require notification of the use of a β2-agonist before an event.

It is a myth that β2-agonists completely prevent EIA in everyone, every time. Most illustrations of the effect of these drugs on EIA use mean data at each timepoint (figure 7), giving the impression of more complete blockade. Analysis of maximum percentage fall in FEV1 reveals incomplete protection in a significant proportion of individuals (figure 8).[6,63]

4.5 LABAs

LABAs can be recommended in patients with normal spirometry who require protection against EIA for <3 days/week.

The advantage of LABAs over SABAs is that they offer longer and more complete protection against EIA for most individuals.[81,82,84,152,155] The onset of the protective effect appears to be no different between the LABAs salmeterol and formoterol against EIA (figure 9).[151] Intermittent usage does not lead to tolerance,[70] and the duration of the protective effect should cover the dosing interval. In the event of breakthrough EIA, a SABA can be used for rescue therapy and is likely to be effective in the standard dose in a person using LABA intermittently.

The disadvantage of using LABAs is that the duration of protection against EIA diminishes with daily use.[67–69] This is true even for patients who only use a single dose of the LABA each day but who also take ICs.[68] This tolerance is likely to be due to downregulation of the β-receptor on mast cells, making them less responsive to the inhibitory effects of a β2-agonist.[60,156] There may be additional downregulation of β-receptors at the level of the airway smooth muscle.[75] Because the tolerance to the protective effect occurs in all people and only a small number of people are required to show significance it is highly unlikely to be genetically based, thus everybody appears to become tolerant. The genetically based problems are different ones with the β-agonists[157] and pertain to people actually getting worse lung function at baseline in response to these drugs. If bronchoconstriction does occur with LABA it may require a higher dose of the SABA than normal for rescue medication.[74,158] In addition, many sporting bodies now require notification of the use of a β2-agonist before an event.

The myths regarding treatment with LABAs are that tolerance does not occur if a person is being treated with ICs[68,77] and that these drugs provide protection from EIA for the entire dose interval.[43,63] In one study, 18% of individuals had a fall in FEV1 >20% only 4 hours after the administration of formoterol.[152] In another study, only 3 of 17 individuals had a <10% fall in FEV1 after exercise 6.5 hours after taking salmeterol 50μg.[63]

In some children with severe EIA, a combination of albuterol and nedocromil has been shown to be more effective than either drug alone,[159] but there do not appear to be any advantages in taking combined therapy (such as a β2-agonist and an anticholinergic agent) for control of EIA.[62] In many countries, oral bronchodilators are still commonly used. Although these drugs are useful for inducing airway smooth muscle relaxation and bronchodilatation, most studies demonstrate efficacy against EIA in only a minority of individuals.[41,160–163]

5. Conclusions

There are many approaches to preventing EIA using single doses of drugs. This review provides information on which factors to base the choice of treatment and the advantages and disadvantages in choosing a particular treatment. Lung function at rest, the severity of EIA and unwanted adverse effects of the drug should be the prime considerations in determining choice. Nowadays many young active people have normal lung function and only require medication to prevent EIA rather than to induce bronchodilatation. Measurements of response to exercise is recommended in order to provide an indirect index of the underlying inflammation and to make an informed choice about the use of ICs. For those asthmatic patients taking β2-agonists on a daily basis, either with or without ICs, consideration should be given to using other drugs to prevent EIA such as cromolyn sodium, nedocromil, and LTRAs. The time has come to make measurements of lung function, to consider the advantages and disadvantages in using the drugs available, and to tailor treatment according to individual needs.