Beta2‐agonists for exercise‐induced asthma
Abstract
Background
It is well known that physical exercise can trigger asthma symptoms and can induce bronchial obstruction in people without clinical asthma. International guidelines on asthma management recommend the use of beta2‐agonists at any stage of the disease. At present, however, no consensus has been reached about the efficacy and safety of beta2‐agonists in the pretreatment of exercise‐induced asthma and exercise‐induced bronchoconstriction. For the purpose of the present review, both of these conditions are referred to by the acronymous EIA, independently from the presence of an underlying chronic clinical disease.
Objectives
To assess the effects of inhaled short‐ and long‐acting beta2‐agonists, compared with placebo, in the pretreatment of children and adults with exercise‐induced asthma (or exercise‐induced bronchoconstriction).
Search methods
Trials were identified by electronic searching of the Cochrane Airways Group Specialised Register of Trials and by handsearching of respiratory journals and meetings. Searches are current as of August 2013.
Selection criteria
We included randomised, double‐blind, placebo‐controlled trials of any study design, published in full text, that assessed the effects of inhaled beta2‐agonists on EIA in adults and children. We excluded studies that did not clearly state diagnostic criteria for EIA.
Data collection and analysis
We used standard methodological procedures as expected by The Cochrane Collaboration.
Main results
We included 53 trials consisting of 1139 participants. Forty‐eight studies used a cross‐over design, and five were performed in accordance with a parallel‐group design. Forty‐five studies addressed the effect of a single beta2‐agonist administration, and eight focused on long‐term treatment. We addressed these two different intervention regimens as different comparisons.
Among primary outcomes for short‐term administration, data on maximum fall in forced expiratory volume in 1 second (FEV1) showed a significant protective effect for both short‐acting beta‐agonists (SABA) and long‐acting beta‐agonists (LABA) compared with placebo, with a mean difference of ‐17.67% (95% confidence interval (CI) ‐19.51% to ‐15.84%, P = 0.00001, 799 participants from 72 studies). The subgroup analysis of studies performed in adults compared with those performed in children showed high heterogeneity confined to children, despite the comparable mean bronchoprotective effect.
Secondary outcomes on other pulmonary function parameters confirmed a more positive and protective effect of beta2‐agonists on EIA compared with placebo. Occurrence of side effects was not significantly different between beta2‐agonists and placebo.
Overall evaluation of the included long‐term studies suggests a beta2‐agonist bronchoprotective effect for the first dose of treatment. However, long‐term use of both SABA and LABA induced the onset of tolerance and decreased the duration of drug effect, even after a short treatment period.
Authors' conclusions
Evidence of low to moderate quality shows that beta2‐agonists, both SABA and LABA, when administered in a single dose, are effective and safe in preventing EIA.
Long‐term regular administration of inhaled beta2‐agonists induces tolerance and lacks sufficient safety data. This finding appears to be of particular clinical relevance in view of the potential for prolonged regular use of beta2‐agonists as monotherapy in the pretreatment of EIA, despite the warnings of drug agencies (FDA, EMA) regarding LABA.
PICOs
Plain language summary
Asthma reliever inhalers (beta2‐agonists) used for exercise‐induced asthma and exercise‐induced bronchoconstriction
Review question
Physical exercise may trigger symptoms such as cough, chest tightness and shortness of breath in people with asthma that is not adequately treated (exercise‐induced asthma). Sometimes people who do not have asthma still experience asthma‐like symptoms during exercise; this is called exercise‐induced bronchoconstriction. We looked at both types of people in this review. The treatments we were interested in are called beta2‐agonists. These are drugs that are known to open up the airways (small tubes in the lungs), making it easier for people to breathe. Two kinds of beta2‐agonists are available: short‐acting (SABA, e.g. salbutamol and terbutaline) and long‐acting (LABA, e.g. formoterol and salmeterol).
What evidence did we find?
We found 53 trials consisting of 1139 participants. Forty‐eight studies used a cross‐over design, which meant that each person in the trial received two or more treatments－one or more active treatments, the beta‐agonist and a placebo in random order. The rest were parallel‐group trials, meaning that people received either the active treatment or a placebo. Most of the studies addressed the effect of a giving a single beta2‐agonist treatment before exercise and recorded the effect on lung function following exercise. Only eight focused on longer treatment－longer treatments would be needed to assess whether these treatments were harmful over the longer term.
Results
Studies in which people received a single administration of a beta‐agonist showed that FEV1 (a measure of lung function) fell significantly less for people taking SABA or LABA compared with placebo (mean difference (MD) ‐17.67%; 95% confidence interval (CI) ‐19.51% to ‐15.84%). Other lung function measures confirmed that beta2‐agonists were more beneficial compared with placebo. No significant difference in the number of side effects was noted in people taking SABA or LABA compared with people taking placebo. However, it is unlikely that people would be prescribed an inhaler for a single treatment, so we must consider longer‐term studies to get a true measure of the side effects that inhalers can cause.
We found that included longer‐term studies showed that beta2‐agonists were helpful in terms of lung function for the first dose of treatment. However, studies that provided longer‐term treatment with SABA or LABA showed that over time, people built up a tolerance to the effects of treatments, and the beneficial effects lasted for shorter periods of time.
Quality of the evidence
Overall, we believe that the evidence was of low to moderate quality.
Conclusions
This review shows that beta2‐agonists－both SABA and LABA－when administered in a single dose, are effective and safe in preventing the symptoms of EIA. Longer‐term administration of inhaled beta2‐agonists induces tolerance and lacks sufficient safety data. It is important to note that taking LABA without background inhaled steroids is considered unsafe and is not currently recommended in most of the clinical guidelines for asthma. We recommend that more studies are needed to determine whether it is safe to administer inhaled beta2‐agonists alone to people who experience asthma symptoms when exercising.
This review is current as of August 2013.
Authors' conclusions
Summary of findings
| Beta2‐agonists compared with placebo (single administration) for exercise‐induced asthma | ||||||
| Patient or population: exercise‐induced asthma | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Placebo (single administration) | Beta2‐agonists | |||||
| Maximal percentage fall in FEV1 | The mean fall in FEV1 in the intervention group was MD 17.67 lower (19.51 lower to 15.84 lower)a | － | 799 | ⊕⊕⊕⊝ | The results in the subgroup of LABA and SABA were similar: MD 15.6 lower (18.29 lower to 12.92 lower) and MD 18.99 lower (21.38 lower to 16.6 lower) in 44 and 28 studies, respectively | |
| Number of participants with an FEV1 fall > 10% | 843 per 1000 (84.3)% | 300 per 1000 | OR 0.08 (0.06 to 0.13) | 773 | ⊕⊕⊕⊝ | |
| Maximal percentage fall in PEF | The mean maximal percentage fall in PEF in the intervention group was MD 24.61 lower (37.57 lower to 11.65 lower)1 | － | 92 | ⊕⊕⊝⊝ | ||
| Maximal percentage fall in FEF25‐75% | The mean maximal percentage fall in FEF25‐75% in the intervention group was MD 20.75 lower (27.17 lower to 14.32 lower)1 | － | 106 | ⊕⊕⊝⊝ | ||
| Side effects | 50 per 1000 (5.0)% | 42 per 1000 | OR 0.83 (0.43 to 1.59) | 2165 | ⊕⊕⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence. | ||||||
| aLower indicates that beta2‐agonists are better than placebo. bIn 51 studies that provided data for subgroup analysis, no difference was observed in the maximal percentage fall in FEV1, but the heterogeneity of the effect was seen primarily in the paediatric population. cThese represent 72 study arms from 53 studies. dIt is unclear how directly pulmonary function measures relate to what participants feel. eThere was concern about lack of concealment, loss to follow‐up and reporting bias. fInconsistency was moderate to high and was explained in part by subgroup analyses of adults and children. gSmall numbers of participants were included with resulting wide confidence intervals. hThese represent 55 study arms rather than studies. iThis represents a mix of outcomes, and not all of them are of equal importance to patients. | ||||||
Background
Exercise‐induced asthma (EIA) is the term commonly used to describe the transient increase in airway resistance that follows vigorous exercise (Anderson 1997). However a recent position paper suggested a preferred definition of exercise‐induced bronchoconstriction (EIB) with or without asthma, depending on the presence of underlying clinical asthma (Weiler 2010). Despite this new nomenclature, most of the relevant studies were published before these terms were proposed and often do not provide sufficient information to distinguish between the two conditions. We therefore decided to adopt the term exercise‐induced asthma (EIA) throughout the review for the sake of better consistency and clarity.
The prevalence of EIA ranges from 5% to 20% in the general population, to even 100% in people with uncontrolled asthma. This huge variability depends not only on the criteria used for diagnosis, but also on the population samples studied. EIA is, in fact, reported to be particularly frequent in children (Randolph 2008), in people with rhinitis (Brozek 2010) and in athletes, with percentages varying according to different sport disciplines (Carlsen 2008).
The diagnosis of EIA is usually made by exercise testing, either in the field or in the laboratory; this second option allows more standardised procedures (Rundell 2000). An individual's response to exercise is generally expressed by the maximal percent fall in forced expiratory volume in one second (FEV1). The maximal percent fall is considered an expression of severity of EIA and is calculated by subtracting the lowest FEV1 value from the pre‐exercise value and expressing it as a percentage of the pre‐exercise value. Both European Respiratory Society (ERS) and American Respiratory Society (ATS) recommendations set a fall threshold of 10% as a diagnostic criterion for EIA and a value greater than 30% as a marker of severe bronchial hyperreactivity, particularly if the person is treated with inhaled steroids (Sterk 1993). Other indirect tests such as eucapnic voluntary hyperpnoea (EVH) and mannitol challenge are usually considered surrogate tests for the diagnosis of EIA because they induce similar pathophysiological changes in the airways (Anderson 2003).
The main principle of treating EIA involves reversing the bronchial obstruction induced by exercise with bronchodilators or preventing it with daily use of either controller drugs in people with asthma (Koh 2007; Bateman 2008) or drugs that inhibit symptoms and improve pulmonary function immediately before exercise. Pretreatment before exercise includes mast cell stabilisers (Kelly 2000; Spooner 2003), leukotriene antagonists (Peroni 2011), short‐acting beta2‐agonists (SABA) and, more recently, long‐acting beta2‐agonists (LABA), especially in endurance athletes (Shapiro 2002).
Both SABA and LABA, administered at standard doses immediately before exercise, have been shown to reduce the fall in FEV1 by 70% to 80% in most people with EIA (Anderson 2006). The mechanism of this protection is believed to be related to beta2‐receptor－induced relaxation of bronchial smooth muscle, which opposes the contractile effects of the various mediators of bronchoconstriction. Protection from EIA is also afforded by beta2‐receptor－induced inhibition of mediator release from mast cells.
At present, however, no consensus has been reached about the efficacy and safety of beta2‐agonists in the pretreatment of EIA. The role of these molecules in preventing EIA was questioned when patients taking beta2‐agonists daily reported breakthrough EIA within a dosing period. Several negative findings have been reported regarding the efficacy of daily treatment with beta2‐agonists in controlling the severity of bronchoconstriction and recovery from EIA. In fact, in a significant minority of people, EIA is not prevented by beta2‐agonists administered at the recommended dose (Anderson 1991; Weiler 2005). Furthermore, it has been reported how daily treatment with beta2‐agonists can enhance the severity of EIA (Hancox 2002) and decrease the duration of their protective effect, especially for LABA (Ramage 1994). In addition, recovery from EIA after a standard dose of beta2‐agonists is slower, and additional doses are often required when SABA or LABA are used daily (Hancox 2002).
On the other hand, the reported association between administration of LABA, not in combination with inhaled corticosteroids, and increased numbers of severe cardiovascular side effects and sudden deaths (Nelson 2006; Salpeter 2010) induced the U.S. Food and Drug Administration (FDA) to set a "black box" on these drugs, highlighting the urgent need to promote clear studies of pharmacovigilance (Martinez 2005).
Objectives
To assess the effects of inhaled short‐ and long‐acting beta2‐agonists, compared with placebo, in the pretreatment of children and adults with exercise‐induced asthma (or exercise‐induced bronchoconstriction).
Methods
Criteria for considering studies for this review
Types of studies
We included double‐blind, randomised controlled trials (RCTs) of any study design. Data published in abstract form only were excluded. At least one primary outcome of this systematic review had to be reported for a study to be considered eligible.
Types of participants
We included children and adults (aged 18 years or older) with a clear history of exercise‐induced asthma and/or a positive response to a standardised exercise challenge, defined according to ERS and ATS guidelines as a fall in FEV1 ≥ 10%. Studies that did not clearly state criteria for EIA diagnosis were excluded.
Types of interventions
Eligible interventions included inhaled beta2‐agonists administered, at any dose, as short‐term or long‐term prophylactic treatment before participants underwent a standardised exercise challenge. SABA and LABA had to be administered within a time period before exercise challenge that did not exceed their pharmacological half‐life (arbitrarily set at 1 hour for SABA, and at 12 hours for LABA). For studies with more than one drug arm, only the comparison with placebo was considered. Studies with more than one drug arm that evaluated different beta2‐agonist molecules were considered as separate trials.
Types of outcome measures
Primary outcomes
-
Mean max % fall in FEV1 (100 × (baseline pre‐exercise value － lowest postexercise value)/baseline pre‐exercise value) in people treated with a beta2‐agonist versus mean max % fall in FEV1 in people treated with placebo
-
Mean % protection afforded by beta2‐agonists (% protection = 100 × (max % fall FEV1 placebo － max % fall FEV1 beta2‐agonist)/max % fall FEV1 placebo)
-
Mean area under the curve (AUC) of time course changes in FEV1 after exercise in people treated with a beta2‐agonist versus mean AUC of time course changes in FEV1 after exercise in people treated with placebo
Secondary outcomes
-
Number of people with a max % fall in FEV1 < 10% (complete protection), < 15% and < 20%
-
Changes from baseline in symptom and sign scores
-
Mean max % fall in other pulmonary function parameters (peak expiratory flow (PEF), forced expiratory flow 25–75% (FEF), maximal expiratory flow at 50% (MEF) etc.)
-
Onset of tolerance (considered for long‐term administration studies and in relation to concomitant treatment with inhaled corticosteroids)
-
Outcomes of physical performance
-
Side/adverse effects
Search methods for identification of studies
Electronic searches
Trials were identified using the Cochrane Airways Group Specialised Register of Trials, which is derived from systematic searches of bibliographic databases, including the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE and CINAHL, as well as handsearching of respiratory journals and meeting abstracts (see Appendix 1 for additional details). The Register was searched using the terms in Appendix 2 from the date of inception up to August 2013. No restriction was placed on language of publication.
Searching other resources
We screened reference lists of included studies, recent reviews and textbooks for relevant citations. We contacted authors of unpublished or 'in‐progress' studies and selected manufacturers of beta2‐agonists to identify additional studies. Furthermore, we searched national and international clinical trial websites (www.clinicalstudyresults.org; www.clinicaltrials.gov; www.fda.gov) for additional trials. Personal contacts with colleagues, collaborators and other trialists working in the field of asthma were at last made to identify further potentially relevant studies. No language or publication restrictions were applied to these searches.
Data collection and analysis
Selection of studies
Titles and abstracts of papers identified in the search were reviewed independently by two review authors (MB, CDM), and articles that appeared to fulfil the inclusion criteria were retrieved. From the full text of these papers, two review authors (MB, EC) independently established whether studies met the inclusion criteria. Studies that did not fulfil all of the inclusion criteria were excluded, and reasons for exclusion were reported. The percentage of agreement was recorded, and any disagreement was solved by consensus. If the two review authors did not reach an agreement, a third review author adjudication (MC) was used to resolve disagreements. In case of further uncertainty, study authors were contacted. Review authors were not blinded to authors, journals, results, etc.
Data extraction and management
Data extraction was performed independently by two review authors (MB, EC). Full texts were screened, and bibliographic details, as well as data regarding study design, participants, disease severity, intervention and outcomes, were recorded in predefined forms and entered into RevMan 5.2. All data, numerical calculations and graphic extrapolations were independently confirmed. We did not deal with missing data.
Assessment of risk of bias in included studies
We assessed the risk of bias in included studies as high, low or unclear using The Cochrane Collaboration’s 'Risk of bias' tool (Higgins 2011) and the following headings.
-
Random sequence generation (selection bias).
-
Allocation concealment (selection bias).
-
Blinding of participants and personnel (performance bias).
-
Blinding of outcome assessment (detection bias).
-
Incomplete outcome data (attrition bias).
-
Selective reporting (reporting bias).
-
Other bias.
Discussion or third party adjudication was used to resolve disagreements when necessary.
Measures of treatment effect
Treatment effects were measured as mean differences (MDs) for continuous outcomes and odds ratios (ORs) for dichotomous outcomes.
Unit of analysis issues
Many included studies were of cross‐over design, but many did not report the results of paired t‐tests for continuous outcomes in this review. Some studies provided raw data that allowed a calculation of the correlation between treatment periods on beta2‐agonists and placebo on maximum percentage fall in FEV1 (see Appendix 3 for the raw data). We used the average correlation from these studies (0.36) to impute an appropriate standard error for within‐‐participant differences (as described in Section 16.4.6.4 of the Cochrane Handbook for Systematic Reviews of Interventions; Higgins 2011). We carried out a meta‐analysis of the mean differences and their standard errors for this outcome using the Generic Inverse Variance method in RevMan 5.2.
Dealing with missing data
Missing data on the correlation between results from participants in the cross‐over studies were imputed using the average correlation from studies that reported appropriate raw data. Cross‐over studies provided information on the number of participants in each arm who experienced a given drop in FEV1, but not the number of participants whose FEV1 dropped on both interventions (active treatment and placebo). We were able to calculate marginal odds ratios for these outcomes but could not adjust the standard error to take advantage of the cross‐over design.
Assessment of heterogeneity
To assess the level of heterogeneity, the Chi2 test and the I2 statistic were used. In establishing the level of heterogeneity, we considered the following rules for interpretation of results (Higgins 2011).
-
0 to 30% as low heterogeneity.
-
30% to 60% as moderate heterogeneity worthy of investigation.
-
60% to 90% as severe heterogeneity worthy of understanding.
-
90% to 100% as allowing aggregation only with major caution.
Assessment of reporting biases
Funnel plots were used to investigate the possibility of publication bias.
Data synthesis
Data were entered into RevMan 5.2. For continuous measures, individual and pooled statistics were reported as mean difference (MD) of treatment effect with 95% confidence intervals (95% CIs) using the fixed‐effect model.
Subgroup analysis and investigation of heterogeneity
Heterogeneity worthy of investigation was examined using the following predefined subgroup analyses.
-
Type of beta2‐agonist (SABA vs LABA).
-
Molecule of beta2‐agonist (formoterol vs salmeterol).
-
Age of participants (children vs adults).
-
Concomitant treatments (beta2‐agonist monotherapy vs concomitant inhaled corticosteroid treatment).
Results
Description of studies
See:Characteristics of included studies; Characteristics of excluded studies.
Results of the search
The search identified 2400 articles, and from these, 289 papers were independently selected by two review authors (k = 0.95) as being of potential interest on the basis of titles and abstracts. After non‐ interventional studies (n = 27) and articles published as abstract only (n = 52) were removed, we retrieved the remaining articles in full text and from these identified 51 studies for inclusion in the meta‐analysis (Figure 1). Two further articles were added through the cross‐checking reference process. We excluded 158 studies with reasons provided in Characteristics of excluded studies; one study is listed under Studies awaiting classification to be considered for inclusion when the review is next updated.
Study flow diagram.
Included studies
Full details can be found in the Characteristics of included studies tables.
Study features and design
All 53 included studies were double‐blind, placebo‐controlled, randomised trials. Forty‐eight studies used a cross‐over design, and five were performed in a parallel ‐group design. In the cross‐over studies, washout periods ranged between 1 and 21 days. Nine studies did not mention duration of washout, and in two papers, no washout was performed. The exercise challenges involved treadmill (n = 35), cycle ergometer (n = 13) and free running (n = 5). All challenges were standardised and met recommended testing criteria. Trials were conducted between 1976 and 2010 in 12 different countries: Europe (N = 27), United States/Canada (N = 22) and Australia (N = 4). All articles except one (in Spanish) were written in English.
Population
Collectively, included studies reported data on 1139 participants. Population sample size ranged from 10 to 161 participants (55 in a single drug arm of parallel‐group studies and 46 in cross‐over studies－the highest number of enrolled participants). Studies included children and adults (age range 4 to 64 years). A total of 20 studies were performed in children, 18 in adults and 12 in both children and adults. Three papers did not provide sufficient information to allocate people according to age. Nine studies provided only information about ethnicity, and Caucasian was the most represented race.
Interventions
Of 53 included studies, 45 addressed beta2‐agonist short‐term administration, and eight focused on long‐term treatment. Articles were grouped on the basis of the type of beta2‐agonist drugs evaluated: SABA (N = 42; short‐term administration n = 40 and long‐term administration n = 2) and LABA (N = 27, short‐term administration n = 21 and long‐term administration n = 6). Among different beta2‐agonists, salbutamol (n = 27), salmeterol (n = 14), formoterol (n = 13) and terbutaline (n = 6) represented the molecules most frequently investigated. Beta2‐agonists were delivered through different devices (nebulisers, metered‐dose inhalers (MDIs) and inhalers).
Details on dosage and types of beta2‐agonist administration are summarised in Table 1, together with timing in relation to exercise.
| Study | Intervention (single dose) | Study type (single‐dose test or chronic [duration]) | When administered before challenge/exercise (min, h) |
| Salbutamol diskus 200 mcg Salbutamol MDI | Single dose | 30 min | |
| Albuterol 180 mcg Salmeterol diskus 25 mcg Salmeterol diskus 50 mcg | Single dose | 30 min, 5 h 30, 11 h 30 | |
| Salbutamol 200 mcg Formoterol 12 mcg | Single dose | 3 h, 12 h | |
| Salbutamol 200 mcg | Single dose | 30 min | |
| Albuterol MDI 180 mcg Albuterol rotacaps 200 mcg | Single dose | 15 min | |
| Salmetrol discus 50 mcg Salmeterol diskhaler 50 mcg | Single dose | 30 min, 5 h 30, 11 h 30 | |
| Albuterol 180 mcg Formoterol 12 mcg Formoterol 24 mcg | Single dose | 15 min, 4 h, 8 h, 12 h | |
| Salmeterol diskhaler 25 mcg Salmeterol diskhaler 50 mcg | Single dose | 10‐12 h | |
| Salbutamol MDI 200 mcg Salbutamol jet disposable 200 mcg | Single dose | 10 min | |
| Fenoterol 100 mcg | Single dose | 10 min | |
| Salbutamol 400 mcg Formoterol 12 mcg | Single dose | 3 h, 12 h | |
| Reproterol 1 mg | Single dose | 15 min | |
| Salmeterol 25 mcg Salmeterol 50 mcg | Single dose | 1 h, 2 h | |
| Salbutamol 200 mcg | Single dose | 20 min | |
| Salbutamol MDI 200 mcg Salbutamol jet device 200 mcg | Single dose | 10 min | |
| Terbutaline 500 mcg | Single dose | 15 min | |
| Terbutaline 250 mcg | Single dose | 1 h | |
| Formoterol 12 mcg | Single dose | 15 min, 4 h | |
| Formoterol 12 mcg twice daily | Long‐term (4 weeks) | 30 min, 12 h at days 1, 14 and 28 | |
| Salmeterol 50 mcg | Single dose | 1 h, 5 h, 9 h | |
| Terbutaline 500 mcg Formoterol 9 mcg Formoterol 4.5 mcg | Single dose | 15 min, 4 h, 8 h | |
| Salbutamol 800 mcg daily | Long‐term (1 week) | 8 h | |
| Salbutamol 180 HFA Salbutamol 180 mcg MDI | Single dose | 30 min | |
| Terbutaline 32.5 mcg | Single dose | 15 min | |
| Salbutamol 200 mcg Formoterol 12 mcg | Single dose | 30 min, 3 h, 5 h 30, 8 h | |
| Salbutamol 200 mcg Salmefamol 200 mcg | Single dose | 20 min | |
| Salbutamol 800 mcg daily | Long‐term (81 weeks) | 24 h | |
| Salbutamol 180 mcg Salmeterol 42 mcg | Single dose | 30 min, 5 h 30, 11 h 30 | |
| Metaproterenol 130 mcg | Single dose | 10 min, 1 h | |
| Fenoterol 400 mcg | Single dose | 10 min | |
| Salbutamol 200 mcg Formoterol 12 mcg | Single dose | 2 h, 4 h | |
| Salbutamol 200 mcg | Single dose | 15 min | |
| Salbutamol 180 mcg | Single dose | 15 min | |
| Rimeterol 400 mcg | Single dose | 2 min | |
| Salmeterol 84 mcg daily | Long‐term (29 days) | 30 min, 9 h | |
| Salbutamol 200 mcg Salmeterol 50 mcg | Single dose | 1 h, 6 h, 12 h | |
| Salbutamol 200 mcg Tolobuterol 200 mcg Tolobuterol 400 mcg | Single dose | 20 min | |
| Salbutamol 200 mcg Formoterol 24 mcg | Single dose | 2 h, 8 h | |
| Salbutamol 180 mcg Formoterol 12 mcg Formoterol 24 mcg | Single dose | 15 min, 4 h, 8 h, 12 h | |
| Salbutamol 90 mcg | Single dose | 20 min | |
| Salmeterol 50 mcg | Single dose | 2 h, 8 h 30, 24 h | |
| Salmeterol 100 mcg daily | Long‐term (28 days) | 6 h, 12 h | |
| Terbutaline 500 mcg Formoterol 12 mcg Salmeterol 50 mcg | Single dose | 5 min, 30 min, 1 h | |
| Salbutamol 180 mcg Formoterol 12 mcg Formoterol 24 mcg | Single dose | 15 min, 4 h, 8 h, 12 h | |
| Salmeterol 50 mcg | Long‐term (28 weeks) | 1 h, 9 h | |
| Formoterol 9 mcg daily | Long‐term (28 weeks) | 1 h, 9 h | |
| Salmeterol 100 mcg daily | Long‐term (28 weeks) | 1 h, 9 h | |
| Fenoterol 400 mcg Salbutamol 200 mcg | Single dose | 30 min | |
| Salbutamol 180 mcg | Single dose | 15 min | |
| Salbutamol 400 mcg | Single dose | 15 min | |
| Bitolterol 1050 mcg | Single dose | 45 min | |
| Terbutaline 500 mcg | Single dose | 25 min, 2 h, 4 h, 6 h |
Outcomes
As per inclusion criteria, all 53 included studies reported data on at least one primary outcome of this review.
Excluded studies
Risk of bias in included studies
An assessment of the risk of bias is presented in the Characteristics of included studies tables and is summarised in a risk of bias figure (Figure 2).
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Most of the included studies were judged at unclear risk for selection bias. Lack of information provided on random sequence generation and on allocation concealment may be explained by the high number of papers (40/53) conducted before 2000, when reporting of these risk of bias criteria was less common.
Blinding
Risks of performance and detection bias were minimised by the narrow inclusion criteria adopted, which allow inclusion in the systematic review of only randomised, at least double‐blind, placebo‐controlled trials.
Incomplete outcome data
Three drug arms ( Konig 1984 Fen 0.4; Konig 1984 Fen 0.8; Boulet 1989 Salb) reported incomplete data on outcomes as specified in this systematic review and were therefore assigned a high risk of bias.
Selective reporting
All studies except one (Ferrari 2000 Form 12) were rated as having unclear risk for reporting bias.
Other potential sources of bias
The possibility of publication bias was investigated in the funnel plot shown in Figure 3. The presence of other potential sources of bias was rated as unclear risk because of the scarce information provided.
Funnel plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.1 Maximal percentage fall in FEV1.
Several papers reported data derived from industry‐funded studies.
Effects of interventions
Effects of intervention were separately assessed for short‐term (single administration) and long‐term beta2‐agonist administration.
Short‐term administration
The 45 studies evaluating short‐term beta2‐agonist administration included 77 arms of active treatment (49 SABA and 28 LABA) in comparison with placebo.
Primary outcomes
Data on max % fall in FEV1 were provided by 77 arms. Effects of SABA short‐term administration were evaluated beyond the pharmacological half‐life (1 hour) in five studies, which were therefore excluded. Analysis of the remaining 72 arms, including 799 participants (Figure 4), showed a significant protective effect of beta2‐agonists compared with placebo (MD ‐17.67%, 95% CI ‐19.51% to ‐15.84%; P = 0.00001). In particular, 58 study arms favoured the active treatment, and 14 trial arms reported no significant difference. Heterogeneity was, however, high (I2 = 71%).
Forest plot of comparison: 1 beta2‐agonists versus placebo (single administration), outcome: 1.1 Maximal percentage fall in FEV1.
Mean % protection was 66% (range 29% to 91%). In seven studies, beta2‐agonist administration not only completely protected participants from EIA but also induced a bronchodilator effect compared with baseline values. In only one case (Bronski 1995 Salb Pwd), placebo offered greater protection compared with the active treatment.
Seventeen studies evaluating LABA short‐term administration at different time points showed an FEV1 % fall AUC that favoured the active treatment. Different study designs prevented merging of the data in a unique comprehensive analysis.
Secondary outcomes
Secondary outcomes on pulmonary function parameters confirmed a positive protective effect of beta2‐agonists on EIA compared with placebo. Complete protection from EIA as assessed by numbers of participants with an FEV1 % fall < 10% was evaluated in 19 studies (Analysis 1.2). A significant difference was noted in the number of participants completely protected (OR 0.08, 95% CI 0.06 to 0.13; P = 0.00001). Similar results were obtained for thresholds of FEV1 fall set at 15% (OR 0.06, 95% CI 0.03 to 0.15) and 20% (OR 0.09, 95% CI 0.06 to 0.14) (Analysis 1.3; Analysis 1.4). Max PEF and FEF25‐75 % fall were assessed, respectively, in 14 and in 8 studies (Analysis 1.5; Analysis 1.6). However, statistical analysis for these two outcomes was based on a limited number of trials because dispersion data were lacking in most of the studies considered. Only three arms (Vasquez 1984 Salb 400; Carlsen 1995 Salm 25; Carlsen 1995 Salm 50) reported data on max MEF25‐50 % fall. No study provided information on changes in symptoms and sign scores and effects on physical performance.
As far as they concern secondary outcomes related to safety, side effects were assessed in 55 trials (Figure 5). Among these, 42 arms reported no adverse event for either active or placebo treatment. Analysis of the remaining 13 trials showed no significant difference between beta2‐agonists and placebo.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.1 Side effects.
Subgroup analysis
The high heterogeneity (I2 = 71%) found for the primary outcome max FEV1 % fall was investigated through the preplanned subgroup analysis.
We performed a subgroup analysis according to types of beta2‐agonists (SABA vs LABA; Figure 6). Although subgroup analysis confirmed a significant bronchoprotective effect against EIA for both classes compared with placebo, it was not able to explain the marked heterogeneity observed in the entire population sample: SABA (I2 = 71%) and LABA (I2 = 67%). Accordingly, non‐significant differences emerged from the analysis of the different beta2‐agonist molecules administered (Figure 7). This evaluation was plotted only for the comparison between formoterol and salmeterol, because the number of studies evaluating different SABA molecules, apart from salbutamol, appeared to be too small for a reliable investigation. It is interesting to note that analysis of studies performed only in adults (n = 19) compared with those performed only in children (n = 32) showed that high heterogeneity was largely confined to studies in children (I2 = 11% and 80%, respectively), despite the comparable mean bronchoprotective effect (Figure 8). Furthermore all studies that failed to show a positive protective effect against EIA of beta2‐agonists compared with placebo dealt with the paediatric population.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.8 Subgroup analysis: maximal percentage fall in FEV1 SABA versus LABA.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.9 Subgroup analysis: maximal percentage fall in FEV1: salmeterol versus formoterol.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.10 Subgroup analysis: maximal percentage fall in FEV1: adults versus children.
The different study designs used and the limited information provided prevented assessment of the potential role of concomitant treatment with inhaled corticosteroids.
Long‐term administration
Long‐term beta2‐agonist administration was addressed in only eight papers. Five trials were performed with a cross‐over design in a total of 64 people (50 adults and 14 children). Three studies adopted a parallel‐group design and included 69 people (49 adults and 20 children) in the active arms and 73 (53 adults and 20 children) placebo controls. Treatment periods range from 7 to 29 days. Effects of SABA were evaluated in two protocols, and LABA administration was assessed in six studies. The limited number of trials, the small population samples and the different study designs and drugs tested allow only a descriptive approach and prevent plotting of data in a meta‐analysis.
Garcia and coauthors (Garcia 2001 Form 12) evaluated the effects of 28‐day formoterol administration in a parallel‐group design (10 people in the beta2‐agonist arm and 9 people in the placebo arm). The protective effect of salmeterol 12 mcg twice daily was assessed on the 1st, 14th and 28th study days, and results were not significantly greater than those provided by placebo at any time point. Furthermore, tachyphylaxis to the beta2‐agonist effect was developed already after two weeks of treatment, although it was not progressive.
Hancox (Hancox 2002) studied the effect of 200 mcg once daily of salbutamol in eight adults treated for seven days in a cross‐over study. Results showed not only an increased max FEV1 % fall after the exercise challenge, but also a sub‐optimal bronchodilator response to further beta2‐agonist administration at the end of the treatment period in the active group.
Ten adults who inhaled 200 mcg of salbutamol or placebo four times a day for seven days were studied by Inman and O'Byrne (Inman 1996) in a cross‐over study. One week of regular inhaled salbutamol resulted in worsening of EIA.
A cross‐over design was also adopted by Nelson (Nelson 1998) in 20 adults treated for a month with inhaled salmeterol 42 mcg twice a day. Significant beta2‐agonist protection against EIA was maintained for the entire study period. However, the length of time that the drug remained active after a single dose significantly decreased. Furthermore, the number of participants for whom salmeterol did not offer complete protection against EIA (FEV1 % fall < 10%) increased from two on study day 1 to 11 on day 29 (P = 0.02)
Ramage and coworkers (Ramage 1994) studied 12 adults treated with inhaled salmeterol 50 mcg twice daily for 4 weeks in a cross‐over manner. The significant protection provided by the first dose of salmeterol against EIA at 6 and 12 hours was no longer present at the end of the treatment period.
Fourteen children were studied by Simons (Simons 1997) in a four‐week cross‐over study. The first dose of salmeterol had an excellent bronchoprotective effect against EIA at 1 and 9 hours. At the end of the study, however, the bronchoprotective effect was significantly greater than that of placebo only at 1 hour.
A parallel‐group study design was adopted by Stelmach (Stelmach 2008) to compare the effects of formoterol 9 mcg daily against placebo in two arms, each consisting of 20 children. Both groups of people were receiving concomitant inhaled corticosteroid (ICS) treatment with budesonide 100 mcg daily. At the end of the four‐week treatment period, bronchoconstriction induced by a standardised treadmill exercise challenge was significantly diminished in the active group compared with the placebo group,
At last, Storms (Storms 2004) in a four‐week parallel‐group study compared the protective effect of salmeterol 50 mcg twice daily against placebo. All enrolled people were receiving concomitant 100 mcg twice‐daily fluticasone treatment. The protective effect against EIA was evaluated at weeks 1 and 4 and was not different between the two groups.
Only one study (Simons 1997) took into consideration safety aspects of long‐term beta2‐adrenergic administration. Reported side effects were minor, were poorly related to study drug and anyway were not different between active and placebo groups.
The overall evaluation of presented studies seems to confirm the beta2‐agonist bronchoprotective effect for the first dose of treatment. However, long‐term use of both SABA and LABA induced the development of tolerance and decreased the duration of drug effect, even after short‐term treatment. The few available data on concomitant therapy with inhaled corticosteroids did not allow a firm statement about their potential influence on the response to beta2‐agonists.
Discussion
Summary of main results
Evidence emerging from the meta‐analysis of 45 short‐term (single administration) studies shows that both short‐ and long‐acting beta2‐agonists administered as preventive treatment (within the time‐effect period set at one hour for SABA and at 12 hours for LABA) prevent exercise‐induced asthma, as shown by the primary outcomes related to the FEV1 fall. This pharmacological effect appears to be clinically relevant and independent of the exercise challenge adopted (treadmill, cycle ergometer, free run). The assessment of secondary outcomes considered shows that the beta2‐agonist preventive effect is also documented by the number of participants protected (complete protection detectable in 68% of participants) and by other pulmonary function variables (PEF, FEF 25%‐75%, MEF 50%) and beta2‐agonists did not cause side effects.
Overall completeness and applicability of evidence
The choice to include only double‐blind randomised trials may influence the completeness of the present review but was thought to reinforce the quality of evidence.
Quality of the evidence
The quality of the evidence gathered for the primary outcome of maximal percentage fall in FEV1 and the number of participants with a fall in FEV1 greater than 10% were moderate owing to some concerns about risk of bias (unclear allocation concealment) and detected and not completely explained inconsistency and indirectness (relation of FEV1 to patient‐important outcomes). The same concerns apply to all other outcomes reported in the 'Summary of findings' (SoF) Table. In addition, for the outcomes of maximal percentage fall in PEF, maximal percentage fall in FEF 25‐75 and side effects, the quality was lowered further by the small numbers of participants for these outcomes and the resulting wide confidence intervals. It can be argued that this review started with the premise that pulmonary function measures are patient‐important outcomes and that, therefore, the quality of evidence should not be lowered for indirectness because the primary outcome was measured in these studies. However, despite these considerations, the meaning of the degree of change in FEV1 for patients remains unclear (as well as how it relates to their well‐being). Furthermore, (small) concerns about inconsistency and risk of bias justify an overall rating of quality as presented in the SoF Table. Further research should focus on patient‐important outcomes and the imprecision that was encountered for many of the secondary outcomes.
Potential biases in the review process
The review process was protected from bias by adherence to a prepublished protocol. We tried to prevent bias in our search process by using comprehensive search terms and by asking study authors to identify other published and non‐published studies. We minimised bias by assessing studies independently and resolving differences of opinion by discussion. Extraction of data and assessment of risk of bias were performed in duplicate as well. We performed only subgroup analyses that were specified a priori in the protocol.
Agreements and disagreements with other studies or reviews
Our results appear to be in agreement with those obtained by Spooner and coworkers in a Cochrane review and meta‐analysis published in 2009 and focused on quantitative comparison of the effects of inhaling mast cell stabilisers (nedocromil sodium or sodium cromoglycate) versus beta2‐agonists (Spooner 2003). However, the review authors confined their review to single‐dose administration and to short‐acting beta2‐agonist molecules. As far as we know, no recent systematic reviews have specifically addressed the efficacy and safety of both SABA and LABA, in short‐term and long‐term administration, for prevention of exercise‐induced asthma.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Funnel plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.1 Maximal percentage fall in FEV1.
Forest plot of comparison: 1 beta2‐agonists versus placebo (single administration), outcome: 1.1 Maximal percentage fall in FEV1.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.1 Side effects.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.8 Subgroup analysis: maximal percentage fall in FEV1 SABA versus LABA.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.9 Subgroup analysis: maximal percentage fall in FEV1: salmeterol versus formoterol.
Forest plot of comparison: 1 Beta2‐agonists versus placebo (single administration), outcome: 1.10 Subgroup analysis: maximal percentage fall in FEV1: adults versus children.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 1 Maximal percentage fall in FEV1.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 2 Number of participants with an FEV1 fall > 10%.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 3 Number of participants with an FEV1 fall > 15%.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 4 Number of participants with an FEV1 fall > 20%.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 5 Maximal percentage fall in PEF.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 6 Maximal percentage fall in FEF 25‐75.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 7 Side effects.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 8 Subgroup analysis: maximal percentage fall in FEV1 SABA vs LABA.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 9 Subgroup analysis: maximal percentage fall in FEV1: salmeterol versus formoterol.
Comparison 1 Beta2‐agonists versus placebo (single administration), Outcome 10 Subgroup analysis: maximal percentage fall in FEV1: adults versus children.
| Beta2‐agonists compared with placebo (single administration) for exercise‐induced asthma | ||||||
| Patient or population: exercise‐induced asthma | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No. of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Placebo (single administration) | Beta2‐agonists | |||||
| Maximal percentage fall in FEV1 | The mean fall in FEV1 in the intervention group was MD 17.67 lower (19.51 lower to 15.84 lower)a | － | 799 | ⊕⊕⊕⊝ | The results in the subgroup of LABA and SABA were similar: MD 15.6 lower (18.29 lower to 12.92 lower) and MD 18.99 lower (21.38 lower to 16.6 lower) in 44 and 28 studies, respectively | |
| Number of participants with an FEV1 fall > 10% | 843 per 1000 (84.3)% | 300 per 1000 | OR 0.08 (0.06 to 0.13) | 773 | ⊕⊕⊕⊝ | |
| Maximal percentage fall in PEF | The mean maximal percentage fall in PEF in the intervention group was MD 24.61 lower (37.57 lower to 11.65 lower)1 | － | 92 | ⊕⊕⊝⊝ | ||
| Maximal percentage fall in FEF25‐75% | The mean maximal percentage fall in FEF25‐75% in the intervention group was MD 20.75 lower (27.17 lower to 14.32 lower)1 | － | 106 | ⊕⊕⊝⊝ | ||
| Side effects | 50 per 1000 (5.0)% | 42 per 1000 | OR 0.83 (0.43 to 1.59) | 2165 | ⊕⊕⊝⊝ | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence. | ||||||
| aLower indicates that beta2‐agonists are better than placebo. bIn 51 studies that provided data for subgroup analysis, no difference was observed in the maximal percentage fall in FEV1, but the heterogeneity of the effect was seen primarily in the paediatric population. cThese represent 72 study arms from 53 studies. dIt is unclear how directly pulmonary function measures relate to what participants feel. eThere was concern about lack of concealment, loss to follow‐up and reporting bias. fInconsistency was moderate to high and was explained in part by subgroup analyses of adults and children. gSmall numbers of participants were included with resulting wide confidence intervals. hThese represent 55 study arms rather than studies. iThis represents a mix of outcomes, and not all of them are of equal importance to patients. | ||||||
| Study | Intervention (single dose) | Study type (single‐dose test or chronic [duration]) | When administered before challenge/exercise (min, h) |
| Salbutamol diskus 200 mcg Salbutamol MDI | Single dose | 30 min | |
| Albuterol 180 mcg Salmeterol diskus 25 mcg Salmeterol diskus 50 mcg | Single dose | 30 min, 5 h 30, 11 h 30 | |
| Salbutamol 200 mcg Formoterol 12 mcg | Single dose | 3 h, 12 h | |
| Salbutamol 200 mcg | Single dose | 30 min | |
| Albuterol MDI 180 mcg Albuterol rotacaps 200 mcg | Single dose | 15 min | |
| Salmetrol discus 50 mcg Salmeterol diskhaler 50 mcg | Single dose | 30 min, 5 h 30, 11 h 30 | |
| Albuterol 180 mcg Formoterol 12 mcg Formoterol 24 mcg | Single dose | 15 min, 4 h, 8 h, 12 h | |
| Salmeterol diskhaler 25 mcg Salmeterol diskhaler 50 mcg | Single dose | 10‐12 h | |
| Salbutamol MDI 200 mcg Salbutamol jet disposable 200 mcg | Single dose | 10 min | |
| Fenoterol 100 mcg | Single dose | 10 min | |
| Salbutamol 400 mcg Formoterol 12 mcg | Single dose | 3 h, 12 h | |
| Reproterol 1 mg | Single dose | 15 min | |
| Salmeterol 25 mcg Salmeterol 50 mcg | Single dose | 1 h, 2 h | |
| Salbutamol 200 mcg | Single dose | 20 min | |
| Salbutamol MDI 200 mcg Salbutamol jet device 200 mcg | Single dose | 10 min | |
| Terbutaline 500 mcg | Single dose | 15 min | |
| Terbutaline 250 mcg | Single dose | 1 h | |
| Formoterol 12 mcg | Single dose | 15 min, 4 h | |
| Formoterol 12 mcg twice daily | Long‐term (4 weeks) | 30 min, 12 h at days 1, 14 and 28 | |
| Salmeterol 50 mcg | Single dose | 1 h, 5 h, 9 h | |
| Terbutaline 500 mcg Formoterol 9 mcg Formoterol 4.5 mcg | Single dose | 15 min, 4 h, 8 h | |
| Salbutamol 800 mcg daily | Long‐term (1 week) | 8 h | |
| Salbutamol 180 HFA Salbutamol 180 mcg MDI | Single dose | 30 min | |
| Terbutaline 32.5 mcg | Single dose | 15 min | |
| Salbutamol 200 mcg Formoterol 12 mcg | Single dose | 30 min, 3 h, 5 h 30, 8 h | |
| Salbutamol 200 mcg Salmefamol 200 mcg | Single dose | 20 min | |
| Salbutamol 800 mcg daily | Long‐term (81 weeks) | 24 h | |
| Salbutamol 180 mcg Salmeterol 42 mcg | Single dose | 30 min, 5 h 30, 11 h 30 | |
| Metaproterenol 130 mcg | Single dose | 10 min, 1 h | |
| Fenoterol 400 mcg | Single dose | 10 min | |
| Salbutamol 200 mcg Formoterol 12 mcg | Single dose | 2 h, 4 h | |
| Salbutamol 200 mcg | Single dose | 15 min | |
| Salbutamol 180 mcg | Single dose | 15 min | |
| Rimeterol 400 mcg | Single dose | 2 min | |
| Salmeterol 84 mcg daily | Long‐term (29 days) | 30 min, 9 h | |
| Salbutamol 200 mcg Salmeterol 50 mcg | Single dose | 1 h, 6 h, 12 h | |
| Salbutamol 200 mcg Tolobuterol 200 mcg Tolobuterol 400 mcg | Single dose | 20 min | |
| Salbutamol 200 mcg Formoterol 24 mcg | Single dose | 2 h, 8 h | |
| Salbutamol 180 mcg Formoterol 12 mcg Formoterol 24 mcg | Single dose | 15 min, 4 h, 8 h, 12 h | |
| Salbutamol 90 mcg | Single dose | 20 min | |
| Salmeterol 50 mcg | Single dose | 2 h, 8 h 30, 24 h | |
| Salmeterol 100 mcg daily | Long‐term (28 days) | 6 h, 12 h | |
| Terbutaline 500 mcg Formoterol 12 mcg Salmeterol 50 mcg | Single dose | 5 min, 30 min, 1 h | |
| Salbutamol 180 mcg Formoterol 12 mcg Formoterol 24 mcg | Single dose | 15 min, 4 h, 8 h, 12 h | |
| Salmeterol 50 mcg | Long‐term (28 weeks) | 1 h, 9 h | |
| Formoterol 9 mcg daily | Long‐term (28 weeks) | 1 h, 9 h | |
| Salmeterol 100 mcg daily | Long‐term (28 weeks) | 1 h, 9 h | |
| Fenoterol 400 mcg Salbutamol 200 mcg | Single dose | 30 min | |
| Salbutamol 180 mcg | Single dose | 15 min | |
| Salbutamol 400 mcg | Single dose | 15 min | |
| Bitolterol 1050 mcg | Single dose | 45 min | |
| Terbutaline 500 mcg | Single dose | 25 min, 2 h, 4 h, 6 h |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Maximal percentage fall in FEV1 Show forest plot | 72 | 799 | Mean Difference (Random, 95% CI) | ‐17.67 [‐19.51, ‐15.84] |
| 2 Number of participants with an FEV1 fall > 10% Show forest plot | 19 | 773 | Odds Ratio (M‐H, Random, 95% CI) | 0.08 [0.06, 0.13] |
| 3 Number of participants with an FEV1 fall > 15% Show forest plot | 13 | 457 | Odds Ratio (M‐H, Random, 95% CI) | 0.06 [0.03, 0.15] |
| 4 Number of participants with an FEV1 fall > 20% Show forest plot | 25 | 1021 | Odds Ratio (M‐H, Random, 95% CI) | 0.09 [0.06, 0.14] |
| 5 Maximal percentage fall in PEF Show forest plot | 14 | 92 | Mean Difference (Random, 95% CI) | ‐24.61 [‐37.57, ‐11.65] |
| 6 Maximal percentage fall in FEF 25‐75 Show forest plot | 8 | 106 | Mean Difference (Fixed, 95% CI) | ‐20.75 [‐27.17, ‐14.32] |
| 7 Side effects Show forest plot | 55 | 2165 | Odds Ratio (M‐H, Random, 95% CI) | 0.83 [0.43, 1.59] |
| 8 Subgroup analysis: maximal percentage fall in FEV1 SABA vs LABA Show forest plot | 72 | Mean Difference (Random, 95% CI) | ‐17.67 [‐19.51, ‐15.84] | |
| 8.1 SABA | 44 | Mean Difference (Random, 95% CI) | ‐18.99 [‐21.38, ‐16.60] | |
| 8.2 LABA | 28 | Mean Difference (Random, 95% CI) | ‐15.60 [‐18.29, ‐12.92] | |
| 9 Subgroup analysis: maximal percentage fall in FEV1: salmeterol versus formoterol Show forest plot | 28 | Mean Difference (Random, 95% CI) | ‐15.60 [‐18.29, ‐12.92] | |
| 9.1 Salmeterol | 13 | Mean Difference (Random, 95% CI) | ‐12.73 [‐16.10, ‐9.37] | |
| 9.2 Formoterol | 15 | Mean Difference (Random, 95% CI) | ‐18.24 [‐22.15, ‐14.34] | |
| 10 Subgroup analysis: maximal percentage fall in FEV1: adults versus children Show forest plot | 51 | Mean Difference (Random, 95% CI) | ‐16.75 [‐19.12, ‐14.39] | |
| 10.1 Adults | 19 | Mean Difference (Random, 95% CI) | ‐18.77 [‐20.78, ‐16.76] | |
| 10.2 Children | 32 | Mean Difference (Random, 95% CI) | ‐15.32 [‐18.88, ‐11.75] | |
