Statins for asthma

Abstract

Background

Asthma is a common chronic respiratory disease. People with asthma have inflammation of their airways that causes recurrent episodes of wheezing, breathlessness and chest tightness, with or without a cough. Statins possess multiple therapeutic effects, including lowering lipid levels in the blood. Statins are reported to have a potential role as an adjunct treatment in asthma. However, comprehensive evidence of the benefits and harms of using statins is required to facilitate decision making.

Objectives

To assess the benefits and harms of statins as an adjunct therapy for asthma in adults and children.

Search methods

We searched for studies in the Cochrane Airways Trials Register, the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE Ovid SP and Embase Ovid SP, from their inception dates We handsearched the proceedings of major respiratory conferences. We also searched clinical trials registries for completed, ongoing and unpublished studies, and scanned the reference lists of included studies and relevant reviews to identify additional studies.

The search is current to 7 February 2020.

Selection criteria

We included randomised controlled trials (RCTs) with a parallel‐group design that assessed statins for at least 12 weeks' duration. We considered all participants with a clinical diagnosis of asthma to be eligible, regardless of age, sex, disease severity and previous or current treatment. We planned to include studies reported as full text, those published as abstract only, and unpublished data.

Data collection and analysis

Two review authors independently screened and selected the studies, extracted outcome data and intervention characteristics from included studies, and assessed risk of bias according to standard Cochrane methodological procedures. We resolved any disagreement through discussion.

Main results

We found only one trial involving a total of 60 people living with asthma. The trial compared the effect of atorvastatin with a placebo (dummy treatment containing lactose) in treating people with chronic asthma. The trial did not report data for the primary outcomes or adverse events. There was uncertainty about the relative effect on forced expiratory volume in one second (FEV₁) and peak expiratory flow (PEF) in the atorvastatin group compared with the placebo group. The study did not report serious adverse effects for the interventions. The included study had internal discrepancies in its reported data.

Authors' conclusions

The evidence was of very low certainty, so we are unable to draw conclusions about the effectiveness and safety of statins to treat asthma. High‐quality RCTs are needed to assess the effect of statins on people with asthma. Well‐designed multicentre trials with larger samples and longer duration of treatment are required, which assess outcomes such as adverse events, hospital utilisation and costs, to provide better quality evidence. Future studies that include subgroups of obese people with asthma are also required.

Plain language summary

Statins for asthma

Review question

We reviewed the evidence about the effect of statins in people with asthma.

Background

Asthma is a common chronic (persistent) airway disease. Asthma is caused by inflammation in the lungs and bronchoconstriction (tightening of the small tubes in the lungs in response to a trigger such as pollen or cold air). Inflammation and bronchoconstriction cause airflow obstruction (narrowing and blockage). People with asthma experience recurrent attacks of wheezing, difficultly in breathing, and chest tightness, with or without a cough.

Statins are anti‐inflammatory medicines. We wanted to review the evidence for statins because the theory is that their anti‐inflammatory properties will help people with asthma. We wanted to discover whether using statin therapy was better or worse than other alternatives, such as usual care alone or different types of statins (atorvastatin, simvastatin).

Study characteristics

We found only one trial that assessed atorvastatin (a type of statin drug) and placebo (dummy treatment) in 60 adults with asthma over a period of 12 weeks. These people had chronic asthma (persistent airway disease) when they started treatment.

Key results

We found that the statin (atorvastatin) did not result in any clear improvement in lung function in people with asthma, but may be better in asthma control than the dummy drug. However, because there is only one trial, we cannot draw firm conclusions.

The trial did not report on adverse effects or severe adverse effects (harms). We are uncertain whether statins may  have beneficial effects on relieving asthma or whether they may increase the risk of adverse events or non‐serious adverse events.

Quality of the evidence

The evidence was of very low certainty, so we cannot draw firm conclusions about whether statins are helpful and safe to be used by people with asthma.

The evidence is current to 7 February 2020.

Summary of findings

Summary of findings 1. Atorvastatin compared to placebo for asthma.

Atorvastatin compared to placebo for asthma
Patient or population: people with asthma Setting: outpatient Intervention: atorvastatin 40 mg Comparison: placebo
Outcomes	Anticipated absolute effects* (95% CI)		Relative effect (95% CI)	No of participants (studies)	Certainty of the evidence (GRADE)	Comments
	Risk with placebo	Risk with atorvastatin
Asthma exacerbations requiring hospitalisation or a course of steroids	See comment	See comment	not estimable	‐	‐	The study did not assess and report asthma exacerbation requiring hospitalisations or a course of steroids for statin and placebo groups.
Health‐related quality of life	See comment	See comment	not estimable	‐	‐	The study did not assess and report health‐related quality of life for statin and placebo groups.
Lung function: change in morning PEF Follow‐up: 12 weeks	The mean change from baseline in morning PEF was 18.96 L/min	The mean change from baseline in morning PEF was 31.28 L/min	MD 12.32 L/min higher (1.55 lower to 26.19 higher)	60 (1 RCT)	⊕⊝⊝⊝ VERY LOW ab
Lung function: change in FEV1 (pre) Follow‐up: 12 weeks	The mean change in FEV1 (pre) was 0.04 L	The mean change in FEV1 (pre) was 0.171 L	MD 0.13 L higher (0.03 lower to 0.3 higher)	60 (1 RCT)	⊕⊝⊝⊝ VERY LOW ab
Lung function: change in FEV1 (post) Follow‐up: 12 weeks	The mean change in FEV1 (post) was 0.104 L	The mean change in FEV1 (post) was 0.215 L	MD 0.11 L higher (0.03 lower to 0.25 higher)	60 (1 RCT)	⊕⊝⊝⊝ VERY LOWab
Serious adverse events or adverse events	See comment	See comment	not estimable	‐	‐	The study did not assess and report serious adverse events or adverse events for statin and placebo groups.
Asthma control: change in ACQ Follow‐up: 12 weeks	The mean change in ACQ was 0.243 points	The mean change in ACQ was 0.65 points	MD 0.41 points lower (0.73 lower to 0.09 lower)	60 (1 RCT)	⊕⊝⊝⊝ VERY LOW ab	A lower score on the ACQ indicates better asthma control.
*The risk in the intervention group (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). ACQ: asthma control questionnaire; CI: Confidence interval; FEV1: forced expiratory volume in one second; MD: Mean difference; PEF: peak expiratory flow; RCT: randomised controlled trial;
GRADE Working Group grades of evidence High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect.

^aDowngraded by one level for study limitations (risk of selection, detection and attrition bias).

^bDowngraded by two levels for inconsistency and discrepancy of data reported in the tables in the published paper.

Background

Description of the condition

Asthma is a common chronic airway disease, which affects up to 18% of the global population. Asthma can affect people of all ages and is characterised by recurrent episodes of wheezing, breathlessness, chest tightness or cough, with variable expiratory airflow limitation (GINA 2020). According to the World Health Survey of adults in 2002 and 2003, the global prevalences of doctor‐diagnosed asthma, clinical/treated asthma, and wheezing were 4.3%, 4.5% and 8.6%, respectively; there was substantial variation in the prevalence of doctor‐diagnosed asthma amongst the 70 countries studied, which ranged from 0.2% in China to 21.0% in Australia (To 2012). The highest prevalence of all categories of asthma was reported in Australia (21.0% for doctor‐diagnosed asthma, 21.5% for clinical asthma and 27.4% for wheezing symptoms). The Health Survey for England, conducted from 2010 to 2011, reported the annual prevalence of asthma (asthma that was reported by patients and diagnosed and treated by clinicians) to be 9.6%, which was equivalent to six million people (Mukherjee 2016). In Malaysia, the prevalences of doctor‐diagnosed asthma and clinical asthma were 5.2% and 5.5%, respectively (To 2012).

The Global Burden of Disease (GBD) study estimated that 26.2 million (20.5 million to 32.6 million) disability‐adjusted life years (DALYs) (overall disease burden counted as the number of years lost due to ill‐health, disability or early death) were lost to asthma in 2015, which is equivalent to 1% (0.9 to 1.3%) of the global burden in that year (GBD 2017). A survey of five Middle Eastern countries, conducted from 2014 to 2016, reported that approximately half of people with asthma had uncontrolled asthma (44.2%, 367 out of 939 participants) and experienced disruption of their work/school routines (49.3%, 408 out of 827 participants) or their family/home responsibilities (48.4%, 400 out of 826 participants) (Mungan 2018).

In 2011, the cost of treating asthma in the UK alone was estimated to be at least GBP 1.1 billion, 74% of which was incurred for primary care services (60% for prescriptions, 14% for consultations) (Mukherjee 2016). The healthcare cost for each person with asthma was reported to be EUR 8554 (ranging from EUR 7411 to 10,199) per annum in Spain (Melero 2018), and USD 267 in South Korea (Lee 2017). In addition to such costs, asthma has an impact on quality of life and affects well‐being physically (e.g. recurrent symptoms), mentally (e.g. stress related to limited daily activities) and socially (e.g. limits to earning/learning performance) (Australian Centre for Asthma Monitoring 2004; Mungan 2018).

Asthma usually begins early in life, with intermittent symptoms that may become persistent over time in genetically susceptible children with an atopic predisposition (a liability to allergic reaction). If not properly controlled, it may progress from mild/moderate asthma to a severe form of asthma (Spahn 2008). Severe asthma is defined as asthma which requires treatment with high‐dose inhaled corticosteroids (ICS), a second controller (systemic corticosteroids), or both, to prevent it from becoming 'uncontrolled', or which remains 'uncontrolled' despite such treatment (Chung 2014).

The prevalence of severe asthma in adults or children is estimated to be 5% to 10% of the total asthma population (Chanez 2007; Chung 2014), although this is uncertain due to variation in the clinical and epidemiological definitions applied. In a study of 65 Dutch pharmacy databases, 3.6% of adults with asthma were diagnosed as having severe refractory asthma (Hekking 2015). It is recognised that the variation in severity of clinical asthma is due to the consequences of different pathological mechanisms (disease progression). For instance, certain mediators such as interleukins (IL) (IL‐1β, IL‐8) and tumour necrosis factor–alpha associated with T‐helper type 1 (Th1) responses (i.e. inflammation) are seen in more severe asthma, while some mediators (IL‐4, IL‐5, IL‐13) associated with T‐helper type 2 (Th2) responses are typical of allergic asthma. Sputum analysis has shown an increase in the numbers of neutrophils and eosinophils (disease fighting white blood cells) in people with asthma, suggesting that combined granulocyte effects contribute to asthma severity (Hastie 2010). The impact of asthma on quality of life and cost of care has become challenging for individuals, family and society. As such, the "goal of management is for people to be free from symptoms and able to lead a normal, active life" (NICE 2017).

Asthma exacerbation is defined as an acute or subacute worsening of symptoms and lung function from the person's usual status. Inhaled short‐acting beta2‐agonists (SABAs), systemic corticosteroids (either oral or intravenous), supplementary oxygen and inhaled ipratropium bromide are used for management of asthma exacerbations (GINA 2020). In some cases, a single dose of intravenous magnesium sulphate infusion may be beneficial (Kew 2014).

Description of the intervention

The standard treatment for asthma involves a variety of regimens, namely bronchodilators and ICS, or combinations of these. The use of statins may be an option as additional therapy to conventional treatment. Statins are medicines that can help to reduce the level of low‐density lipoprotein (LDL) cholesterol in the blood.

Statins are inhibitors of the rate‐limiting enzyme 3‐hydroxy‐3‐methylglutaryl coenzyme A (HMG‐CoA) reductase in cholesterol biosynthesis (fat synthesis within human body tissues and blood). Statins act primarily via effects on lipids, by inhibiting the intracellular synthesis (synthesis inside the body cells) of L‐mevalonic acid and its metabolites (McKay 2004). The overall cascade of mechanisms in the mevalonate pathway contributes to a reduction in serum cholesterol levels.

At present, there are five types of statins available on prescription in the UK: atorvastatin, fluvastatin, pravastatin, rosuvastatin and simvastatin (NHS UK 2016). In the USA, lovastatin and pitavastatin are also available (FDA 2014). Of these, rosuvastatin and pravastatin are hydrophilic (attracted to water molecules), while lovastatin, simvastatin, fluvastatin and atorvastatin are lipophilic (attracted to fat molecules).

How the intervention might work

A characteristic feature of asthmatic airways is an enhanced inflammatory process. Statins may work through anti‐inflammatory mechanisms. In addition to anti‐inflammatory lipid‐dependent effects, statins have lipid‐independent (pleiotropic) effects. The mechanisms responsible for the anti‐inflammatory effects include:

1. reduction of pro‐inflammatory cytokines by inhibiting mRNA signalling pathways (Inoue 2000);

2. inhibition of lymphocyte proliferation by reduction of cytokine levels (Rudich 1998);

3. production of antioxidant effects through inactivation of reactive oxygen species (ROS) (Wagner 2000);

4. up‐regulation of endothelial nitric oxide synthase expression (Laufs 1998);

5. reduction of bronchial smooth muscle proliferation in response to mitogens (Takeda 2006).

In the context of clinical practice, the incidence of hospitalisation/emergency department visits was reported to be one‐third lower in people with asthma who used statins compared to those who did not (Stanek 2009). However, these findings could be confounded due to factors including age and gender of participants, type of healthcare facility and previous history of asthma therapy. Statin users also had at least two fewer prescriptions of oral corticosteroids compared to non‐statin users in a large cohort study of 14,566 people carried out in the USA (the Population‐based Effectiveness in Asthma and Lung Diseases (PEAL) cohort) (Tse 2013). Additionally, asthma can be associated with obesity (Dixon 2011), and with diabetes (Black 2011). The prevalence of asthma was found to be 10% among youths with type 1 diabetes and 16% among youths with type 2 diabetes (Black 2011). Hence, statin‐containing therapies targeted to such population subgroups may have a double benefit by lowering cholesterol and inducing anti‐inflammatory effects in those with obesity and those with both obesity and diabetes. Thus, statins may have a role as an adjunct treatment of asthma.

Why it is important to do this review

Asthma is responsible for a large number of hospital admissions and presents a substantial burden to emergency departments. In the UK alone, most admissions of people with asthma are emergency cases, and 70% of hospital admissions due to asthma could have been prevented by appropriate early intervention (NICE 2017). Standard preventative therapies for asthma are, however, not without potential risk. Long‐term usage of high‐dose ICS as a controller medication may have local adverse effects in the oral cavity and pharynx (e.g. oral candidiasis, dry throat, hoarseness, thirst and altered sense of taste) (Pinto 2013), as well as systemic side effects (e.g. cataracts, glaucoma, osteoporosis, skin thinning and behavioural abnormalities) (Hanania 1995; Tattersfield 2004). An alternative option for long‐term asthma management is the use of a leukotriene receptor antagonist (LTRA). However, a Cochrane Review reported that the effect of adding LTRAs to a higher dose of ICS was uncertain (Chauhan 2017). Moreover, there had been reports of increased risk of Churg‐Strauss syndrome (a clinical condition that can cause difficulty in breathing, facial pain and a persistent runny nose) with LTRAs (Hauser 2008). Long‐term use of salmeterol monotherapy could also contribute to serious adverse events, such as loss of asthma control, with a subsequent increase in asthma‐related mortality (Cates 2009). Therefore, it is important to implement effective and safe treatment options in order to achieve proper asthma control and to reduce acute exacerbations.

Numerous studies have shown potential therapeutic effects of statins in the management of asthma, due to the anti‐inflammatory properties of these medications. A cross‐sectional study of 165 people with severe asthma, conducted in the USA, reported that statin users achieved better asthma control than non‐users (Zeki 2013). A number of non‐Cochrane reviews on the efficacy of statins in people with asthma, which included both RCTs and non‐RCTs, did not provide conclusive evidence for a clinically significant improvement in participants' outcomes (Silva 2012; Yuan 2012). In addition, there was a risk of negative health effects from the use of statins, mainly muscle and liver effects. A large study on the effect of statins on skeletal muscle function and performance suggested that statins lead to mild muscle injury, even among asymptomatic people (Parker 2013). The US Food and Drug Administration (FDA) also issued a concern over drug interactions, reporting that protease inhibitors (i.e. drugs for HIV or hepatitis C virus) and statins, when taken together, may raise the blood levels of statins and increase the risk of myopathy. Similarly, there were reported cases of simvastatin‐induced rhabdomyolysis (serious skeletal muscle damage) after co‐administration with erythromycin (a strong inhibitor of cytochrome P450 3A4) (Fallah 2013). A systematic screening of the World Health Organization's Adverse Drug Reaction database (VigiAccess) revealed 53 cases of rhabdomyolysis as a result of interaction between statins and azithromycin, of which 12 cases (22.6%) used no drugs other than azithromycin and a statin (Strandell 2009). Though statins are found to present a relatively higher risk of transaminase elevations (elevated liver enzymes in blood) than placebo (Kashani 2006), clinically apparent liver injury with statin therapy is very rare (Mach 2018). Thus, it is important to undertake a comprehensive assessment of all the available data on both the benefits and harms of statins as an adjunct treatment for asthma.

Objectives

To assess the benefits and harms of statins as an adjunct therapy for asthma in adults and children.

Methods

Criteria for considering studies for this review

Types of studies

We included randomised controlled trials (RCTs) with a parallel‐group design, of at least 12 weeks' duration. We included studies reported in full text, as well as those published as an abstract only, and unpublished data.

We did not include studies with a cross‐over design, because some statins (e.g. atorvastatin, rosuvastatin) could have long‐lasting effects.

Types of participants

We included participants with a clinical diagnosis of asthma, regardless of age, sex, disease severity, previous or current treatment.

We excluded studies in which participants with primary or secondary immunodeficiency (e.g. retroviral infection, people on immunosuppressant drugs) were receiving statins.

We had planned to include studies that recruited participants with other pulmonary diseases (e.g. chronic obstructive pulmonary disease (COPD)), provided that we could obtain the results of the subset of participants with asthma separately. However, we did not encounter such studies.

Types of interventions

We included studies that compared statins (any dose and any route of administration) with placebo or different types of statin, in addition to current standard care for asthma. We also included studies that compared statins with usual care and no placebo. Studies were eligible for inclusion if the use of a statin was the only systematic difference between treatment arms. We included co‐interventions if they were not part of the randomised treatment (e.g. inhaled corticosteroids, long‐acting beta₂‐agonists, oral aminophylline).

We included trials that compared .

1. Statin (e.g. atorvastatin) versus placebo

2. Statin versus a different statin

3. Statin versus usual care with no placebo

Types of outcome measures

We assessed the primary and secondary outcomes described below. 

Primary outcomes

1. Participants with exacerbations requiring hospitalisation

2. Participants with exacerbations needing a course of steroids

3. Health‐related quality of life (change from baseline, assessed with validated questionnaire)

4. Serious adverse events (all‐cause mortality, any life threatening events)

Secondary outcomes

1. Adverse events/side effects, e.g. change from baseline in alanine transaminase (ALT) and creatinine kinase, number of participants with rhabdomyolysis

2. Measures of lung function (forced expiratory volume in ml per second (FEV₁))

3. Measures of airway inflammation, e.g. blood eosinophil (count — absolute); sputum or bronchoalveolar lavage eosinophil (%); fractional exhaled nitric oxide (FeNO); C‐reactive protein (CRP)

4. Asthma control measured on a validated scale, e.g. Asthma Control Questionnaire (ACQ) (Juniper 1999) or Asthma Control Test (ACT) (Liu 2007; Nathan 2004).

We extracted data for each outcome relative to background controller therapy, e.g. inhaled corticosteroids (ICS), if this was mentioned .

Search methods for identification of studies

Electronic searches

We identified studies from the following databases and trials registries:

1. Cochrane Airways Trials Register (Cochrane Airways 2019), via the Cochrane Register of Studies, all years to 7 February 2020;

2. Cochrane Central Register of Controlled Trials (CENTRAL), via the Cochrane Register of Studies, all years to 7 February 2020;

3. MEDLINE Ovid SP (ALL) 1946 to 7 February 2020;

4. Embase Ovid SP 1974 to 7 February 2020;

5. US National Institutes of Health Ongoing Trials Register ClinicalTrials.gov (www.clinicaltrials.gov) all years to 7 February 2020;

6. World Health Organization International Clinical Trials Registry Platform (apps.who.int/trialsearch) all years to 7 February 2020.

The database search strategies are listed in Appendix 1. The Cochrane Airways Information Specialist developed the search strategies, in collaboration with the authors.

We searched all databases and trials registries from their inception up to 7 February 2020, and placed no restriction on language or type of publication. We handsearched conference abstracts and grey literature through the Cochrane Airways Trials Register and the CENTRAL database.

Searching other resources

We searched the reference lists of all primary studies and review articles for additional references. We searched relevant manufacturers' web sites for study information.

We searched for errata or retractions from included studies published in full text on PubMed and PubPeeron 16 March 2020.

Data collection and analysis

Selection of studies

Two review authors (CN, HN) screened the titles and abstracts of the search results independently and coded them as 'retrieve' (eligible or potentially eligible/unclear) or 'do not retrieve'. We retrieved the full‐text study reports of all potentially eligible studies, then independently screened them for inclusion, recording the reasons for exclusion of ineligible studies. We resolved any disagreement through discussion. We identified and excluded duplicates, and collated multiple reports of the same study so that each study, rather than each report, was the unit of interest in the review. A PRISMA flow diagram (Figure 1) shows the details of the selection process, and the 'Characteristics of excluded studies' table documents the reasons for studies' exclusion (Moher 2009).

1.

Study flow diagram.

Data extraction and management

We used a data collection form for study characteristics and outcome data, which had been piloted on at least one study in the review. One review author (CN) extracted the following study characteristics from included studies.

1. Methods: study design, total duration of study, details of any 'run‐in' period, number of study centres and location, study setting, withdrawals and date of study.

2. Participants: number (n), mean age, age range, gender, severity of condition, diagnostic criteria, baseline lung function, smoking history, inclusion criteria and exclusion criteria.

3. Interventions: intervention, comparison, concomitant medications and excluded medications.

4. Outcomes: primary and secondary outcomes specified and collected, and time points reported.

5. Notes: funding for studies and notable conflicts of interest of trial authors.

Two review authors (CN, HN) independently extracted outcome data from included studies. We noted in the 'Characteristics of included studies' table if the study did not report outcome data in a usable way. We resolved disagreements by consensus. One review author (CN) transferred data into the Review Manager 5 file (RevMan 2014). We double‐checked that data were entered correctly by comparing the data presented in the systematic review with the study reports. A second review author (HN) spot‐checked study characteristics for accuracy against the study report. If we deemed the studies eligible but needed clarification, we contacted the author of the primary study to obtain additional information.

Assessment of risk of bias in included studies

Two review authors (CN, HN) assessed risk of bias independently for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2017). We resolved any disagreements by discussion. We assessed the risk of bias according to the following domains:

1. random sequence generation;

2. allocation concealment;

3. blinding of participants and personnel;

4. blinding of outcome assessment;

5. incomplete outcome data;

6. selective outcome reporting;

7. other bias.

We judged each potential source of bias as high, low or unclear, and provided a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We summarised the 'Risk of bias' judgements across studies for each of the domains listed. We considered blinding separately for different key outcomes where necessary (e.g. for unblinded outcome assessment, risk of bias for all‐cause mortality might be very different to that for a participant‐reported pain scale). Where information on risk of bias related to unpublished data or correspondence with a trialist, we noted this in the 'Risk of bias' table. When considering treatment effects, we took into account the risk of bias for the studies that contributed to that outcome.

Assessment of bias in conducting the systematic review

We conducted the present review according to the published protocol (Naing 2019), and justified any deviations from it in the 'Differences between protocol and review' section of the systematic review.

Measures of treatment effect

For dichotomous data (e.g. presence/absence of adverse events), we planned to report the odds ratio (OR) with 95% confidence interval (CI). For continuous data (e.g. FEV₁, PEF), we used the mean difference (MD). We could have used standardised mean difference (SMD) if trials used different scales. If data from rating scales were to have been combined in a meta‐analysis (i.e. ACQ, ACT), we would have ensured that they were entered with a consistent direction of effect (e.g. lower scores indicate improvement). We planned to undertake meta‐analyses when the studies' treatments, participants, underlying clinical questions and time points were similar enough for pooling to make sense.

We planned to describe skewed data narratively (e.g. medians and interquartile ranges for each group).

Where multiple trial arms were reported in a single study, we planned to include only the relevant arms. If two comparisons (e.g. atorvastatin versus placebo and simvastatin versus placebo) were combined in the same trial, we could have either combined the active arms or halved the control group to avoid double‐counting. If both change‐from‐baseline and endpoint scores were available for continuous data, we used change‐from‐baseline scores. If a study reported outcomes at multiple time points, we planned to use the data collected at the end of the study. However, the included trial had no such issue.

We used intention‐to‐treat (ITT) or 'full analysis set' analyses where they were reported (i.e. those where data had been imputed for participants who were randomly assigned but did not complete the study), instead of per protocol analyses.

Unit of analysis issues

For dichotomous outcomes, we planned to use participants, rather than events as the unit of analysis. For example, the number of participants with adverse events, rather than the same participants frequently reported to a particular form of adverse event, or the number of children admitted to hospital, rather than number of admissions per child. However, the included trial had no such issue.

Dealing with missing data

We contacted the original author (Dr. Bardar) to clarify missing data, but we did not receive any response.

Assessment of heterogeneity

We planned to measure heterogeneity among the studies in each analysis using the I² statistic. If we identified substantial heterogeneity, we had planned to report it and explore the possible causes by prespecified subgroup analysis. As we could identify only one trial, this was not required.

Assessment of reporting biases

We planned to create a funnel plot to explore possible small‐study and publication biases if we could identify more than ten studies. As we could identify only one trial, we did not carry out this assessment.

Data synthesis

We converted the mean and standard error of the mean (SEM) data expressed in the original report into mean and standard deviation, and then reported the data in a narrative manner. As there was only one included study, we were unable to pool the treatment effects.

To ensure robustness of the analysis, we had planned to pool the data using the random‐effects model and then carry out sensitivity analysis with a fixed‐effect model. We could identify only one trial (Badar 2014), in which there was discrepancy of data related to the outcome measurements in the published report. Due to a lack of additional information required from the primary author, we did not create any forest plot.

Subgroup analysis and investigation of heterogeneity

We planned to carry out subgroup analyses for the following:

1. age group (adults, children);

2. severity of asthma (severe asthma, non‐severe asthma as defined by European Respiratory Society/American Thoracic Society (ERS/ATS) guidelines) (Chung 2014);

3. different statins (e.g. atorvastatin, simvastatin);

4. co‐interventions (e.g. ICS, long‐acting beta agonists, oral aminophylline).

We planned to use the following outcomes in subgroup analyses:

1. asthma exacerbations requiring hospitalisation;

2. asthma exacerbations needing a course of steroids;

3. serious adverse events;

4. measures of lung function.

We had planned to use the formal test for subgroup interactions in Review Manager 5 (RevMan 2014).

Sensitivity analysis

We planned to perform a sensitivity analysis by removing studies with high or unclear risk of bias from the meta‐analyses. Therefore, the analysis would have only included studies at low risk of bias in all key domains, namely, adequate generation of the randomisation sequence, adequate allocation concealment and blinding of outcome assessor for the estimates of treatment effect. We also planned to perform a sensitivity analysis by comparing the results from inclusion and exclusion of the imputed data values.

Summary of findings and assessment of the certainty of the evidence

We created a 'Summary of findings' table using the outcomes listed below. We used the five GRADE considerations (risk of bias, consistency of effect, imprecision, indirectness and publication bias) to assess the quality of a body of evidence as it relates to the studies that contribute data for the prespecified outcomes. We used the methods and recommendations described in Section 8.5 and Chapter 12 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019; McKenzie 2019), using GRADEpro GDT software (GRADEpro GDT). We justified all decisions to downgrade the quality of studies using footnotes, and made comments to aid the reader's understanding of the review where necessary.

We included the following outcomes in the 'Summary of findings' table:

1. asthma exacerbations requiring hospitalisation;

2. asthma exacerbations needing a course of steroids;

3. health‐related quality of life;

4. all reported serious adverse events;

5. measures of lung function;

6. all reported adverse events;

7. asthma control (ACQ or ACT).

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies

Results of the search

The search generated 270 records, and we screened 217 records after removing duplicates. We excluded 196 records after screening titles/abstracts, and reviewed the full texts of 21 studies. We included only one study that met the review's inclusion criteria, and excluded the remaining 20 studies (Figure 1). In order to ensure that we had identified any eligible ongoing studies, we contacted an expert in this field and also checked clinical trials registries. We did not find any ongoing trials that measure the effectiveness of statins on asthma treatment outcomes.

Included studies

Badar 2014 is the only included study in this review. It involved people with chronic asthma who attended the Chest and Tuberculosis outpatient department of the tertiary care government hospital in central India, during the period from June 2012 to December 2012. The trial enrolled 60 participants, randomised into two groups (30 in each group). The trial compared oral atorvastatin 40 mg with matching placebo (an inactive substance, which consisted of lactose). The majority of participants were men in the atorvastatin group (56.7%) with a mean age of 43.1(± 10.6) years. In the placebo group, the participants were mainly women (60%), with a mean age of 25.7 (± 18.4) years. Mean duration of asthma was 24.1 years for the atorvastatin group and 25.7 years for the placebo group.

Excluded studies

We excluded 20 studies from this review. The main reasons for exclusion were;

• not a specified 12‐week duration (seven studies: Braganza 2011; Fahimia 2009; Maneechotesuwan 2010; Mohammad 2012; Moini 2012; Sun 2017; Thomson 2015);

• cross‐over design (three studies: Cowan 2010; Hothersall 2008; Menzies 2007);

• not on people with asthma (three studies: Albert 2011; Gehin 2015; Keddissi 2007);

• not an RCT (five studies: Huang 2015; Ostroukhova 2009; Tse 2013; Wang 2018; Zeki 2013);

• review only (two studies: Silva 2012; Yuan 2012).

Risk of bias in included studies

See Characteristics of included studies; Figure 2 for detailed assessment.

2.

Risk of bias summary: review authors' judgements about each risk of bias item for the included study.

Allocation

Due to unclear information in the published article about the method of randomisation and whether or not allocation was concealed, we classified Badar 2014 as having an unclear risk of selection bias.

Blinding

Badar 2014 had a low risk of performance bias, as the trial report described it as "a double‐blind placebo controlled clinical trial" and mentioned the use of matched placebo. However, it had an unclear risk of detection bias as the trial report did not provide information about who performed the outcome assessments.

Incomplete outcome data

We assessed Badar 2014 to have a high risk of attrition bias, as the withdrawal rate was high and somewhat unbalanced (10% in the treatment arm and 6.7% for the controls).

Selective reporting

Since we could not identify the protocol of this study, we judged Badar 2014 to have an unclear risk of reporting bias.

Other potential sources of bias

Badar 2014 had an unclear risk of other bias. We could not clarify the accuracy of data in the published report, despite our requests for additional information from the study author.

Effects of interventions

See: Table 1

Table 1 presents the key results for the included study.

Only one trial reported the outcomes of interest, measured at 12 weeks' duration of treatment (Badar 2014).

Primary outcomes

Badar 2014 did not report the following outcomes that we had planned for this review.

1. Participants with exacerbations requiring hospitalisation

2. Participants with exacerbations needing a course of steroids

3. Health‐related quality of life (change from baseline, assessed with validated questionnaire)

4. Serious adverse events (all‐cause mortality, any life threatening events).

Secondary outcomes

Badar 2014 reported the outcomes PEF, FEV₁ (pre), FEV₁ (post), ACQ and sputum cell counts. We created forest plots using data from Table 3 in the published article. However, we noticed that the data in Tables 2 and 3 in the article are discrepant. We contacted the investigators for clarification, but received no reply.

PEF

The mean change from baseline for morning PEF was 31.28 ± 6.26 L/min in the atorvastatin group and 18.96 ± 3.3 L/min in the placebo group (P > 0.05). Very low‐certainty evidence showed a mean difference in morning PEF between atorvastatin and placebo groups of 12.32 L/min (95% CI ‐1.55 to 26.19; Analysis 1.1).

1.1. Analysis.

Comparison 1: Atorvastatin vs placebo, Outcome 1: Change from baseline in morning PEF (L/min)

FEV₁ (pre)

The mean change from baseline for FEV₁ (pre) was 0.171 ± 0.44 L in the atorvastatin group and 0.04 ± 0.12 L in the placebo group (P > 0.05). Very low‐certainty evidence showed a mean difference in FEV₁ (pre) between atorvastatin and placebo groups of 0.13 L (95% CI ‐0.03 to 0.30; Analysis 1.2).

FEV₁ (post)

The mean change from baseline for FEV₁ (post) was 0.215 ± 0.32 L in the atorvastatin group and 0.104 ± 0.22 L in the placebo group (P > 0.05). Very low‐certainty evidence showed a mean difference in FEV₁ (post) between atorvastatin and placebo groups of 0.11 L (95% CI ‐0.03 to 0.25; Analysis 1.3).

1.3. Analysis.

Comparison 1: Atorvastatin vs placebo, Outcome 3: Change in FEV₁ (post) L/min

ACQ

The mean change from baseline in ACQ score was ‐0.65 ± 0.71 in the atorvastatin group and ‐0.24 ± 0.54 in placebo group (P < 0.05). Very low‐certainty evidence showed a mean difference in asthma control between atorvastatin and placebo group of ‐0.41 (95% CI ‐0.73 to ‐0.09; Analysis 1.4). On this scale, a higher score indicates greater impairment.

1.4. Analysis.

Comparison 1: Atorvastatin vs placebo, Outcome 4: Change in ACQ

Sputum cell counts

The mean change from baseline for sputum cell counts (x10⁶) was ‐0.41 ± 0.4 in the atorvastatin group and ‐0.11 ± 0.24 in the placebo group (P < 0.05). Very low‐certainty evidence showed a mean difference in reduction of sputum cell counts between atorvastatin and placebo groups of ‐0.30 (95% CI ‐0.47 to ‐0.13; Analysis 1.5).

1.5. Analysis.

Comparison 1: Atorvastatin vs placebo, Outcome 5: Change in sputum cell count (10⁶)

Badar 2014 did not report the following outcomes that we had planned for this review.

1. Adverse events/side effects,

2. Measures of airway inflammation such as FeNO and CRP.

Discussion

Summary of main results

The findings of this review were based on one randomised, double‐blind, placebo controlled trial involving a total of 60 participants with chronic asthma. We could not perform any meta‐analyses.

Our systematic review has demonstrated the following.

1. A single trial studied the effects of atorvastatin and presented the selected outcomes measured at 12 weeks' duration of treatment (Badar 2014).

2. There were uncertain differences between atorvastatin and placebo in terms of FEV₁ (pre), FEV₁ (post) and morning PEF.

3. There were differences between atorvastatin and placebo in ACQ and sputum cell counts. However, this was subject to discrepancies of data in the published report.

4. We were not able to assess the main primary outcomes of this review.

5. We were not able to include subgroup analyses due to limited data.

Overall completeness and applicability of evidence

On comprehensive checking of 217 abstracts screened in the literature search, we could identify only one small RCT that compared atorvastatin and matching placebo for asthma at 12 weeks' duration of treatment (Badar 2014). Our multiple attempts to contact the author for further information on the trial, and to clarify the data through email and phone contact, were unsuccessful.

We also contacted the investigators of potential studies in order to verify key study characteristics or to obtain missing/additional numerical outcome data (Braganza 2011; Maneechotesuwan 2010; Moini 2012; Sun 2017; Thomson 2015). All except one study author (Professor Thomson) provided additional information. We excluded these potential trials mainly because they assessed the effect of statins on people with asthma at time points shorter than the specified minimum 12‐week duration of treatment for this review (Braganza 2011; Fahimia 2009; Maneechotesuwan 2010; Moini 2012; Sun 2017; Thomson 2015).

Findings from this review are applicable to adults with asthma. However, as the evidence is based on a small trial, caution is required when interpreting the results. The study included was conducted in India, which is a lower middle‐income country; therefore, the role of statin use for asthma in high‐income countries is not known. In addition, we could not find RCTs that reported outcomes such as cost and quality of life, which will guide clinicians' treatment decisions.

Moreover, there have been reports on the issue of muscle‐related side effects of statins by The European Atherosclerosis Society (Stroes 2015), and by the American Heart Association (Newman 2019). However, the only included study in our review did not report the adverse events of statins in participants with asthma.

Certainty of the evidence

We rated the certainty of the evidence of this review was as very low, as we could not perform meta‐analyses to synthesise the overall treatment effects with only one included RCT. Furthermore, we assessed the study to be at high risk of attrition bias and at unclear risk of selection, detection and reporting biases. Moreover, we detected an obvious discrepancy of data in Table 2 and Table 3 in the published report, for which we could not obtain clarification from the study author.

There were several limitations to this systematic review that need to be acknowledged. Firstly, as the results are based on one small study it could reduce the evaluation power and increase the potential for publication bias in this review. Secondly, the results may have been affected by the methodological quality of the primary study and lack of consistency of data in the primary report. In addition, there was a gender imbalance in participants of the two groups, with more women than men in the placebo group. A nationally representative survey conducted in the US reported that, compared with non‐obese men, obese men were more likely to have asthma, with a 1.3‐fold increase after adjustment for common factors (OR 1.28, 95% CI 1.11 to 1.49) (Wang 2015). A subgroup analysis in a Cochrane Review documented that the effect of atorvastatin was greater in women than in men (Adams 2015). These suggest that the asthma‐obesity association is gender‐dependent. Moreover, 13.3% of participants in the atorvastatin group and 10% in the placebo group in the included trial were former smokers. Hence, hormonal effect, smoking status, or a combination of these in participants might play a role in the effect estimates. Finally, the present review may have selective reporting bias. We could not identify a protocol for the included study, and the publication did not report data on clinically important outcomes (e.g. asthma exacerbations). There was also a lack of data on health‐related quality of life and treatment‐related outcomes such as adverse events.

Potential biases in the review process

This review was conducted according to a published protocol (Naing 2019), which aims to reduce bias in review conduct. Different inclusion/methodological decisions may have given a different result, for example we only included trials over 12 weeks.

Agreements and disagreements with other studies or reviews

The results of this Cochrane Review reflect the very low‐certainty evidence on the benefits and harm of statins for people with asthma following a 12‐week duration of treatment. There is a non‐Cochrane systematic review that consisted of five trials (both parallel and cross‐over designs), comprising 200 participants with asthma and with durations of 4 to 12 weeks. It also reported insufficient evidence to show whether statins improve lung function in people with asthma in terms of FEV₁ and morning PEF (Si 2013). Similarly, the results of a cross‐sectional retrospective study of obese adults with severe asthma (31 statin users and 134 non‐users) indicated that statin use was associated with better asthma control (in terms of ACT scores), but no differences in lung function (in terms of FEV₁, PEF) or peripheral eosinophilia (Zeki 2013). These findings raised concerns that there might be a subpopulation (subphenotype) of asthma that can benefit from statins. Obese people, for example, are more refractory to standard inhaler treatment. Sun 2017 reported that obesity becomes a special phenotype of asthma in clinic, and asthma in obese people is different from that in non‐obese people, with more severe clinical manifestations.

There is a Cochrane Review which reported that statins for people with COPD had a reduction in IL‐6 and CRP than the placebo in people with COPD with no clear clinical benefit. Due to the small numbers of included studies and participants, the evidence provided was, however, of low quality (Walsh 2019). In addition, the characteristics of COPD and asthma are not identical.

Authors' conclusions

Implications for practice.

The evidence presented in this review is insufficient to permit recommendations regarding the effectiveness or safety of statins as an adjuvant therapy to standard treatment. Based upon data from only one randomised controlled trial (RCT), atorvastatin did have advantages over placebo in terms of Asthma Control Questionnaire (ACQ) score and reduction of sputum cell counts. Nevertheless, such evidence should be weighed in the context of the published report, as well as patient‐related factors in the clinical setting.

Implications for research.

Further large‐scale, well‐designed multi‐centre RCTs with a sufficiently long duration of treatment should be conducted, that address patient‐related outcomes. Pair‐wise (head‐to‐head) comparisons using various form of statins may provide information on how to treat asthma most effectively. There are several phenotypes in asthma with poor prognosis. Future research should include people with asthma who have different phenotypes, to assess the robustness of potential effects of statins. Since currently available data revealed no study on utilisation and cost‐effectiveness of statins, future research on these outcomes is required.

History

Protocol first published: Issue 2, 2019
Review first published: Issue 7, 2020

Appendices

Appendix 1. Database search strategies

Cochrane Airways Trials Register & CENTRAL (via the Cochrane Register of Studies)

#1 AST:MISC1
#2 MeSH DESCRIPTOR Asthma Explode All
#3 asthma*:ti,ab
#4 #1 or #2 or #3
#5 MeSH DESCRIPTOR Hydroxymethylglutaryl‐CoA Reductase Inhibitors
#6 statin*
#7 atorvastatin
#8 fluvastatin
#9 lovastatin
#10 pitavastatin
#11 pravastatin
#12 rosuvastatin
#13 simvastatin
#14 #5 or #6 or #7 or #8 or #9 or #10 or #11 or #12 or #13
#15 #4 AND #14

MEDLINE Ovid SP

1. exp asthma/
2. asthma$.tw.
3. 1 or 2
4. exp Hydroxymethylglutaryl‐CoA Reductase Inhibitors/
5. (statin$ or atorvastatin or fluvastatin or lovastatin or pitavastatin or pravastatin or rosuvastatin or simvastatin).tw.
6. 4 or 5
7. 3 and 6
8. (controlled clinical trial or randomized controlled trial).pt.
9. (randomized or randomised).ab,ti.
10. placebo.ab,ti.
11. dt.fs.
12. randomly.ab,ti.
13. trial.ab,ti.
14. groups.ab,ti.
15. or/8‐14
16. Animals/
17. Humans/
18. 16 not (16 and 17)
19. 15 not 18
20. 7 and 19

Embase Ovid SP

1. exp asthma/
2. asthma$.tw.
3. 1 or 2
4. exp hydroxymethylglutaryl coenzyme A reductase inhibitor/
5. (statin$ or atorvastatin or fluvastatin or lovastatin or pitavastatin or pravastatin or rosuvastatin or simvastatin).tw.
6. 4 or 5
7. 3 and 6
8. Randomized Controlled Trial/
9. randomization/
10. controlled clinical trial/
11. Double Blind Procedure/
12. Single Blind Procedure/
13. Crossover Procedure/
14. (clinica$ adj3 trial$).tw.
15. ((singl$ or doubl$ or trebl$ or tripl$) adj3 (mask$ or blind$ or method$)).tw.
16. exp Placebo/
17. placebo$.ti,ab.
18. random$.ti,ab.
19. ((control$ or prospectiv$) adj3 (trial$ or method$ or stud$)).tw.
20. (crossover$ or cross‐over$).ti,ab.
21. or/8‐20
22. exp animals/ or exp invertebrate/ or animal experiment/ or animal model/ or animal tissue/ or animal cell/ or nonhuman/
23. human/ or normal human/ or human cell/
24. 22 and 23
25. 22 not 24
26. 21 not 25
27. 7 and 26

ClincalTrials.gov

Search field	Search terms
Study type	Interventional
Condition	asthma
Intervention	statin OR atorvastatin OR fluvastatin OR lovastatin OR pitavastatin OR pravastatin OR rosuvastatin OR simvastatin

WHO portal

Search field	Search terms
Condition	asthma
Intervention	statin OR atorvastatin OR fluvastatin OR lovastatin OR pitavastatin OR pravastatin OR rosuvastatin OR simvastatin

Data and analyses

Comparison 1. Atorvastatin vs placebo.

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1.1 Change from baseline in morning PEF (L/min)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.2 Change in FEV1 (pre) L/min	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.3 Change in FEV1 (post) L/min	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.4 Change in ACQ	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected
1.5 Change in sputum cell count (106)	1		Mean Difference (IV, Fixed, 95% CI)	Totals not selected

1.2. Analysis.

Comparison 1: Atorvastatin vs placebo, Outcome 2: Change in FEV₁ (pre) L/min

Characteristics of studies

Characteristics of included studies [ordered by study ID]

Badar 2014.

Study characteristics
Methods	Design: randomised, double‐blind, parallel placebo‐controlled trial Total duration: 12 weeks Blinding: double Setting: outpatient Location: Chest and TB department of the hospital in Central India Date of study: June 2012 to December 2012 Withdrawals: 2
Participants	Number screened: not mentioned Number randomised: 69 Lost to follow up: 7, withdrawals: 2 due to adverse events and asthma exacerbation Allocation: atorvastatin 40 mg (n = 30), placebo (n = 30) Numbers completing trial: atorvastatin 40 mg (n = 30), placebo (n = 30) Age in years, mean (±SD): atorvastatin, 43.13 (10.61); placebo, 41.73 (11.41) Males (%): atorvastatin (56.7), placebo (40) Duration of asthma in years (±SD): atorvastatin 24.1 (15), placebo 25.7(18.4) Daily puffs of reliever (salbutamol), mean (±SD): atorvastatin 2.1(2.1), placebo 2.5 (2) Ex‐smokers: atorvastatin = 4, placebo = 3 Inclusion criteria: aged 18 to 70 years; chronic asthma diagnosed with spirometry and clinical examinations; duration of asthma: ≥ 1 year; stable medication for 4 weeks. Exclusion criteria: history of medication on a statin or medications known to interact with statins; known sensitivity to statin or myopathy; pregnant or lactating women; abnormal creatinine kinase or liver function tests; and with myositis.
Interventions	atorvastatin 40 mg daily matched placebo (lactose) Comparison: atorvastatin (40 mg per day) versus placebo for 12 weeks Inhaled salbutamol and beclomethasone ‐ both treatment arms for a further 4 weeks.
Outcomes	measures of lung function: FEV1 (pre), PEF, FVC, asthma control questionnaire (ACQ) asthma exacerbations sputum cell count
Notes	Funding: none Identifier: none
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Described as "60 patients randomly divided in 2 groups", with no further details.
Allocation concealment (selection bias)	Unclear risk	Information not available in published article.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Described as "a double‐blind placebo controlled clinical trial" with the use of matched placebo.
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Unclear who performed the assessments.
Incomplete outcome data (attrition bias) All outcomes	High risk	Withdrawal rate was high (10% in the treatment arm and 6.7% in the placebo group).
Selective reporting (reporting bias)	Unclear risk	Not known, no protocol identified.
Other bias	High risk	Inconsistency of data to compute the effect size.

SD: standard deviation; TB: tuberculosis

Characteristics of excluded studies [ordered by study ID]

Study	Reason for exclusion
Albert 2011	Participants do not have asthma.
Braganza 2011	Study duration is eight weeks.
Cowan 2010	A randomised, placebo‐controlled, cross‐over trial
Fahimia 2009	Study duration is four weeks.
Gehin 2015	A single‐centre, open‐label study in healthy male subjects
Hothersall 2008	A randomised, placebo‐controlled, cross‐over trial
Huang 2015	Not a randomised trial
Keddissi 2007	Participants do not have asthma
Maneechotesuwan 2010	Study duration is eight weeks
Menzies 2007	A randomised, double‐blind cross‐over trial
Mohammad 2012	Study duration is eight weeks
Moini 2012	Study duration is eight weeks
Ostroukhova 2009	Not a randomised trial
Silva 2012	A review, not a randomised trial
Sun 2017	Study duration is 55 days
Thomson 2015	Study duration is eight weeks
Tse 2013	Cohort study
Wang 2018	Not a randomised trial
Yuan 2012	A review
Zeki 2013	Not a randomised trial

Differences between protocol and review

The data synthesis methods described in the protocol could not be implemented in the current review because we only included one trial. 

Contributions of authors

Cho Naing (CN) developed the protocol with suggestions and input from Han Ni (HN).

Both authors (CN and HN) contributed to conducting the full review, and agreed the version prior to submission for editorial review.

Contributions of editorial team

Chris Cates (Co‐ordinating Editor) checked the data entry prior to the full write up of the review; edited the review; advised on methodology; and approved the review prior to publication.

Emma Dennett (Managing Editor): co‐ordinated the editorial process; advised on interpretation and content; edited the review.

Emma Jackson (Assistant Managing Editor): conducted peer review and edited the plain language summary and reference sections of the review.

Elizabeth Stovold (Information Specialist): designed the search strategy; ran the searches; and edited the search methods section.

Sources of support

Internal sources

• The authors declare that no such funding was received for this systematic review, Other

External sources

• The authors declare that no such funding was received for this systematic review, Other

Declarations of interest

Cho Naing: none known.

Han Ni: none known.

New