This paper reviews a number of clinical trials and outcome studies on the use of acupuncture to treat asthma. None of the trials had a large sample size and they suffer from the problems common to all acupuncture trials, such as what constitutes appropriate treatment and a suitable 'control'. The outcome studies avoid some of these problems but many lack rigour or adequate description of measurements used.
This paper reviews the trials that appeared most sound in methodological terms, together with those outcome studies that had relatively large samples. The trials are divided into two groups: those treating patients as part of normal clinical management and those treating patients suffering an induced asthma attack. The findings for the first group are inconsistent, particularly as regards objective measures of lung function. There is, however, evidence that acupuncture can improve patients' subjective experience of their symptoms, reduce their use of medication and improve immunological parameters. The smaller number of trials of induced asthma is more consistently positive. The majority of trials provide an inadequate rationale for the acupuncture points used and few bear any relationship to the way acupuncture is actually practised by British Acupuncture Council members. The outcome studies generally provide a much better picture of the diagnostic criteria used and show more consistently positive results than the trials, but their methodology is often weak. The paper concludes by discussing some of the issues that need to be addressed in developing clinically relevant and methodologically sound research.
The Evidence Series of Briefing Papers aims to provide a review of the key papers in the literature, which provide evidence of the effectiveness of acupuncture in the treatment of specific conditions. The sources of evidence will be clearly identified ranging from clinical trials, outcome studies and case studies. In particular this series of briefing papers will seek to present, discuss and critically evaluate the evidence.
It is estimated that asthma affects approximately 10% of the population. Both incidence of the disease and resulting mortality are increasing (Howell, 2000). The definition of 'asthma' is not, however, clear-cut. Howell (2000) identified three elements: reversible airways obstruction with episodic attacks of breathlessness accompanied by wheezing; responsiveness to asthma drugs (cromoglycate and/or corticosteroids); bronchial hypersensitiveness. In practice, however, none of these provides an absolute criterion for distinguishing asthma from other breathing problems and asthmatics are defined as those with reversible airways obstruction and/or clear responsiveness to asthma medication. The management of asthma often requires daily use of medication on a prophylactic basis. Severe attacks may require hospitalisation and can lead to death. Asthma thus has high economic and personal costs. Lewith and Watkins (1996) suggested that asthma costs the NHS about £400 million per annum; the Department of Social Security, in the form of sickness benefits, £60 million; the economy as a whole, in terms of lost productivity, £350 million.
A search was carried out using the ARRCBASE, the Acupuncture Research Resource Centre database of articles drawn from the British Library's AMED and the US MEDLINE, using the terms 'asthma', 'breathlessness' and 'acupuncture'. Sixty-one references were identified and an additional twenty-two were derived from citations in these papers. Papers were excluded for variety of reasons: no English translation was available; the paper was unobtainable from the British Library; the paper was not primarily concerned with presenting or reviewing clinical evidence; the study involved therapies other than acupuncture or did not use needles. The remainder could be grouped into three categories: reviews, descriptions of a randomised controlled trial (RCT) and descriptions of an outcome study. Reviews were only included if they had been published after 1990, on the grounds that earlier publications would have been superseded. Some RCTs were excluded on the grounds of their weak methodology. For trials prior to 1990, the system of evaluating methodological quality set out by Kleijnen et al (1991) was used to exclude those which scored below 40. Trials since 1990 were included if fully randomised. Outcome studies were included if they used significantly larger sample sizes than the RCTs. The papers finally selected comprise 6 reviews, 11 controlled trials, and four outcome studies.
Two of the reviews (Kleijnen et al (1991) and Linde et al (2000)) focused on the methodological quality of the trials reviewed. Kleijnen et al (1991) reviewed 13 trials, which were evaluated against 18 predefined methodological criteria and scored out of 100. Only 8 studies scored above 50 and no paper scored above 72, leading the authors to conclude that no conclusion as to acupuncture's effectiveness could be drawn due to the poor quality of the trials. A similar conclusion was reached in the most recent review carried out by Linde, Jobst and Panton (Linde et al 2001) as part of the Cochrane Collaboration. This review involved tight selection criteria: of 21 trials identified, only seven were regarded as worthy of inclusion.
The aim of the Cochrane collaboration is to provide overall analyses that can show conclusively whether there is evidence in favour of particular medical interventions. The authors argued that the heterogeneous nature of the acupuncture trials precluded such an analysis. Both these reviews proposed that there was no conclusive evidence that acupuncture has a significant effect on the course of the condition - but equally there is no evidence that it does not. In effect, these reviews are an argument for more rigorous research. Until then, the 'jury is out'. The remaining four reviews (Linde et al (1996), Jobst (1995, 1996), Lewith and Watkins (1996)) analysed a range of individual trials, all of which involved some comparison of a treatment group receiving true acupuncture with a control group receiving sham acupuncture. Table 1 summarises their findings.
These reviews highlight the difficulty in reaching agreement, partly because of the heterogeneous nature of the trials themselves and partly because of differences in the subjective interpretation of the reviewers. For example, Linde et al (1996) used a set of assessors to evaluate both the methodological validity of the trials and the appropriateness of the acupuncture treatments used. Whilst their assessors reached a high level of agreement on the internal validity of the trials, there was little agreement amongst them as to whether the acupuncture treatments given were appropriate; also, the outcome measures used varied considerably.
Table 1: Reviews of Acupuncture
Author Number of
Conclusions Linde et al
14 True acupuncture superior to sham: 5
Trend in favour of true acupuncture: 2
No difference between true and sham: 6
Sham acupuncture superior: 1
16 True acupuncture superior: 10
No difference: 3
10 9 positive on at least some outcome
Furthermore they disagreed with Kleijnen et al (1991) in some of their assessments as to whether or not individual trials showed positive results. Similarly, Jobst (1995,1996) differs from Linde et al (1996) in the interpretation of the results of two trials. In summary, three of the reviews, Kleijnen et al (1991), Linde et al (1996) and Linde et al (2000), argued that the trials did not enable us to come to any conclusions about acupuncture's effectiveness, whilst three, Jobst (1995,1996) and Lewith and Watkins (1996), considered that there was evidence of effectiveness. Jobst (1995,1996) suggested that acupuncture might be used as an addition to conventional medical management of asthma and could lead to a reduction in the need for medication, particularly corticosteroids. Lewith and Watkins (1996) concluded that acupuncture could be useful in the alleviation of short term, acute airways obstruction but that evidence for its long-term efficacy was more open to question - largely, however, because the majority of the trials failed to include any long-term follow-up. The next sections will review in more detail some of the RCTs discussed in the above reviews as well as outcome studies, which, being uncontrolled, are rarely considered in systematic reviews.
- Randomised Controlled Trials
Randomised Controlled Trials
Eleven trials were selected, of which four concern acupuncture administered shortly after the onset of induced asthma, whilst the remainder cover acupuncture provided under normal clinical conditions. Clearly, the latter are of greatest interest to practitioners since most asthmatics encountered in the treatment room will have their asthma managed through drugs and practitioners may very rarely treat a severe acute attack. Of more interest will be issues such as whether acupuncture can lead to a reduction in medication. Nevertheless, the studies of induced asthma are of interest if they can demonstrate whether acupuncture has an effect.
Trials of treatment given under normal clinical conditions.
Characteristics of trials
The trials detailed in Table 2 all involved patients with a diagnosis of chronic asthma, apart from Jobst et al (1986) where the diagnosis was Chronic Obstructive Pulmonary Disease, and only four of the subjects had signs of asthma. This paper was included since it appears in all the reviews of trials of asthma. The outcome measures used in the studies varied. All but Joos et al (2000) included measures of lung function. Other measures included:
- medication use
- immunological parameters
- heart rate and blood pressure
- walking distance (Jobst et al (1986) only)
- subjective relief of symptoms
- subjective well-being, quality of life measures.
The details of the outcome measurements used are included in the footnote to Table 2.
Table 2: Controlled Trials for Chronic Asthma
et al (1984)
17 10 over
Ren 17, LI4,
13 vs sham.
Lung function: MPEFR,
EPEFR Medication: no. of
puffs of "-agonist
Subjective: DSA, WSA
Immunological: IgE, IgG, IgA,
Modest effect of appropriate acupuncture on both
objective and subjective measures of lung function
and one immunological parameter. More substantial
effect on medication use.
Dias et al
20 Variable Ren 22,
Dingchuan, Lu 7
vs GB 5 & 6
Lung function: PEFR
Level of medication usage
Improvements in both groups but control group better
than appropriate acupuncture.
Jobst et al
26 13 over
Lung function: PEFR, FEV1,
Subjective measures of
Walking distance: six minute
No change in lung function. Significant improvement
in well being and walking distance for appropriate
31 8 over 12
Ren 17, Bl 13,
Liv 3 vs Sp 8, Ki
Lung function: PEFR
Asthma symptoms: patient
No. of Asthma episodes
Improvements in both groups. No statistically
significant difference between them. Appropriate
acupuncture group had no asthma episodes compared
with four in control.
25 8 over 4
LI 4, St 36, Du
14, Lu 7,
Lung function: SGaw,
Diaries of medication use &
Heart rate and BP.
Trend to improvement in both groups but not
Ren 17 vs sham
point on the
Lung function: FEV1, FVC.
Quality of life questionnaire.
No improvement in lung function, both groups had
improved quality of life and reduced medication.
Joos et al
Bl 13, 17, Ren
17, LI 4, Lu7
Immunological parameters: 14
General well-being (patient
Significant improvement in general well-being and
most immunological parameters for appropriate
* The inappropriate points also included both a set of basic points for all patients (TE3, 19, GB 8, 34) and randomly assigned flexible points (Bl 38, 55, St 4, 6, 32, TE 14, 23, SI 5). Key to abbreviations: MPEFR (morning peak expiratory flow rate), EPEFR (evening peak expiratory flow rate), PEFR (peak expiratory flow rate), FEV1 (forced expiratory volume in one second), FEF50 or 75 (forced expiratory flow after 50% or 75% vital capacity exhaled), Raw (airway resistance), SGaw (specific airway conductance), DSA (daily severe asthma scale), WSA (weekly severe asthma scale), BP (blood pressure). Methodologically, it is extremely problematic to design a double blind trial in which both patient and practitioner are blinded. If the treatment is provided by a trained practitioner, even if they are given sets of points to needle by a different practitioner, they may be able to identify whether points are inappropriate or appropriate for the condition being treated. Where sham points are used, the problem is insurmountable. In practice, therefore, the trials described in Table 2 as 'double blind' have blinded the patients and used a blinded evaluator but the practitioner providing the treatment is not necessarily blinded. There may therefore be little difference between trials which describe themselves as double or single blind. The majority of the trials involved some sort of period during which baseline measurements of parameters such as lung function were drawn up, followed by a treatment period, followed by further measurements. The majority did not involve any long-term follow-up of patients. Two trials (Biernacki and Peake (1998), Tashkin (1985)) used a crossover design, whereby patients were randomly assigned to real or placebo acupuncture, followed by a washout period, followed by a second treatment phase during which they received the alternate form of acupuncture to the one received in the previous treatment phase. As for the actual treatment given, it is unfortunate that the RCT design has come to be associated with the idea of standard treatments. Whilst this constraint has been more open to question in recent years, only two of the trials below included any individualisation of treatment. In one (Jobst et al (1986)) treatment was fully individualised according to TCM (Traditional Chinese Medicine) syndromes whilst in the other (Joos et al (2000)) both standard and individualised points were used. The control group received either sham acupuncture (points with no defined energetic effect) or what were defined as inappropriate acupuncture points. Researchers differed as to whether they thought the control points should be located reasonably close to the 'real' points or at some distance. All trials except Dias et al (1982) gave a standard number of treatments, which varied from one (real) treatment in Biernacki and Peake (1998) to thirteen in the Jobst et al (1986) study. The number of treatments in the Dias et al (1982) study varied from 2 to 8 (median 6) in the control group and 4 to 12 (median 6) in the treated group.
As regards objective outcome measurements, six of the seven trials measured lung function, and, in four, patients experienced improvements in lung function. However, one of these favoured inappropriate acupuncture over appropriate acupuncture and two failed to show a statistically significant change. Only one, therefore, unequivocally favoured appropriate acupuncture. Two of the trials looked at immunological parameters, both of which demonstrated positive benefits for appropriate acupuncture. Joos et al (2000) reported positive changes in a number of immunological parameters, although only the increase in in vitro lymphocyte proliferation rates reached statistical significance when comparing the TCM group with the control group. Christensen et al (1984) reported reduced levels of IgE in the true acupuncture group. Turning to subjective indicators, six of the trials used measures such as general well-being, quality of life or subjective experience of symptoms. All showed patients experiencing benefits, with three showing appropriate acupuncture superior to inappropriate and three showing improvements in both groups. The relative importance of objective and subjective measures is debated: Jobst et al (1986) argued that acupuncture was helpful in reducing disability since the subjective experience of breathlessness and ability to walk for six minutes improved significantly even without there being a corresponding change in objective measures of lung function. In relation to all these trials we should bear in mind that they had small sample sizes and only two made any attempt to include some element of individual diagnosis. Both of these (Jobst et al (1986), Joos et al (2000)) included positive outcomes. The problem of diagnosis and point choice will be discussed further later.
Trials of acupuncture for induced asthma attacks.
Three trials looked at the effects of acupuncture on people with a history of asthma, but where bronchospasm had been induced, either by exercise (Fung et al (1986), Chow et al (1983)) or by inhalation of methacholine (Tashkin et al (1977)). A fourth (Yu & Lee (1976)) looked at acupuncture as a treatment for spontaneous asthma attacks, but a sub-group of four patients had an attack induced by histamine inhalation whilst in remission. Table 3 summarises the trial characteristics.
Table 3: Controlled Trials for Induced Asthma
16* 1 before
lung area vs
Lung function: FEV1, FVC. Neither gave protection against asthma. Tashkin
12 1 after
St36, Lu7 vs
Lung function: FVC, FEV1,
FEF25-75, FVC, Sgaw, Raw, Vtg.
Real acupuncture better than sham although
isoproterenol was most effective.
19** 1 before
K3 vs SI14, P4,
Lung function: FEV1, FVC,
Real acupuncture gave greater protection than sham. Yu &
20 1 during
and 1 after
St 36 (both
Lung function: FEV1, FVC,
Dingchuan showed significant benefit compared with
both sham and St36 during spontaneous asthma
* all aged 8 - 13. Needles left in during exercise. ** all aged 9 - 13.5 years Abbreviations: as Table 3 + PaCo2 (arterial carbon dioxide pressure) and Vtg (thoracic volume at functional residual capacity).
Three of the four trials showed positive findings regarding relief of bronchospasm by acupuncture, although Tashkin et al (1977) found that medication had a stronger effect. Yu and Lee (1976) found acupuncture to be beneficial in terms of both subjective and objective parameters for spontaneous asthma attacks, but not effective for histamine-induced attacks. This study is of particular interest because of the comparison of the effect of the extra point Dingchuan with both St 36 and a non-acupuncture site. Nine of the 10 patients in the group treated with Dingchuan experienced relief from breathlessness. This was superior to the non- acupuncture site, which was located 4 cms lateral to Dingchuan. The point St 36 was least effective, with only one out of 20 patients reporting any benefit.
Four outcome studies have been included. These are summarised in Table 4.
Three of the outcome studies are particularly useful because they provided for individualised treatments and explained the diagnostic criteria used in the choice of points. The studies all demonstrated very positive results, but a weakness is that the baseline and outcome measures are often unclear. Where specific measures were referred to, actual statistics were rarely given, instead rather vague categories, such as 'markedly improved', were used. An exception is the experimental group in the Shao Jingming study.
Whilst three of the four trials of induced asthma showed the acupuncture conferred statistically significant improvements in objective and subjective symptoms, the findings of the trials for asthma as part of normal clinical practice were more mixed. Nevertheless, they demonstrate evidence for the efficacy of appropriate acupuncture for some immunological parameters, for experience of asthma symptoms, level of medication use and quality of life. Findings are more equivocal for improvements in objective measures of lung function with only one, Christensen et al (1984), finding a statistically significant effect in favour of appropriate acupuncture. One, Dias et al (1982),
Table 4: Outcome Studies
Acupuncture Outcome measures Results Zang
192 Single treatment Lu 6 & Lu 10 with electro acupuncture Clinical observation of symptoms and
signs such as dyspnoea, wheezing but not
clear how measured.
remission or marked
143 Treated for six
Bl 13, 20, 23, Du 14, Ren 15, 22 (all)
Dingchuan or St 40 for excess type. Bl 43, Ren
4 (sometimes with moxa) for deficiency type.
Asthma, chest distress, dyspnoea, cough,
expectoration, wheezing and cyanosis. Not
clear how measured.
cure or markedly
111 Daily then alternate
days over ten days.
Bl 12, 13, Du 14 (all). Lu 5,9 (cough). Ren 12
St 36 (sputum). Bl 23, Ren 4, Ki 3 (kidney
deficiency). Moxa used for cold symptoms,
cupping for heat. Dietary restrictions.
Reference to symptoms but not specified.
Lung function tests in experimental group
9-12 sessions per
Individualised according to eight principles, but
no information on actual points used.
Observed improvements in asthma
symptoms plus various objective
measures: suprarenal and hypophysis
functions, immune status, tryptophan
exchange, physical development.
Positive effect for
87% of all patients.
* A comparison of the effects of Bl 13, Du 14 and Bl 12 using acupuncture with cupping, acupuncture with moxa and no treatment showed the greatest improvement for acupuncture with cupping. They also compared Bl 13, Du 14 and Bl 12 and found that Bl 13 gave the best results. found in favour of the control group, which received 'inappropriate' acupuncture. This leads some researchers, e.g. Grebski et al (1999), to argue that acupuncture is a useful placebo but that the exact location of the needles is unimportant. Three counter-arguments can be made. Firstly, as Joos et al (2000) point out, inappropriate acupuncture is not a placebo since it has definite physiological effects. 'True' acupuncture would, therefore, have to show a greater effect to achieve statistical significance than if it were being compared with an inert placebo. Secondly, a number of trials showed 'real acupuncture' to be superior and the study by Yu and Lee (1976) indicates that, in an acute attack of asthma, it makes a great deal of difference where the needle is inserted, with the point Dingchuan showing a markedly more positive effect than two other points. Thirdly, the majority of the trials used standardised formulae with no attempt at individual diagnosis. Appropriate points included points on the lung, large intestine, kidney, liver and stomach channels together with Ren 17 and 22 and the Back-shu point of the lung, but there was rarely any clear rationale for the choice. Stomach 36, for example, was used by Tashkin et al (1977, 1985) as a 'real point' whereas Yu and Lee (1976) found that it was no better than a sham point. Most traditional acupuncturists would, therefore, regard these trials as of little help in understanding the potential role of acupuncture as it is actually practised. The outcome studies bear a closer relationship to the practice of traditional acupuncture and also demonstrate very positive results. However, they suffer from poor design. Further studies are clearly needed which combine both rigorous research methods and good quality acupuncture treatment. There is no reason why outcome studies cannot use well-validated tools to clearly specify the base line and outcome measures rather than using vague terms such as 'significant improvement'. It is also argued by many that pragmatic RCTs, where the acupuncture treatment is left to the practitioner's discretion and the control group receives an alternative form of treatment or no treatment at all, are more useful than placebo-controlled trials because they enable acupuncture to be studied as it is actually practised. 'We think it more important to know if acupuncture is of value for the patient than to know that it is 'more than placebo'' (Linde et al, 1996).
Biernacki W, Peake MD (1998) Acupuncture in treatment of stable asthma. Resp. Med.
Chow O et al (1983) Effect of acupuncture on exercise induced asthma. Lung, 161:321-6.
Christensen PA et al (1984) Acupuncture and Bronchial Asthma. Allergy 39, 379 - 385
Dias PLR, Subramaniam S, Lionel N D W. (1982) Effects of acupuncture in bronchial
asthma: preliminary communication. J. R. Soc. Med. 75: 245 - 248
Fung KP, Chow OKW, So SY. (1986) Attenuation of exercise induced asthma by
acupuncture. Lancet. December 20-27: 1419-1421.
Grebski E et al (1999) Long-Term Effects of Real and Sham Acupuncture on Lung Function
and Eosiphilic Inflammation in Chronic Allergic Asthma: Randomised, Prospective Study.
European Respiratory Journal 14:507s (abstract only).
Howell J (2000) Asthma: clinical descriptions and definitions in Busse W, Holgate S Asthma
and rhinitus Vol 1. Oxford, Blackwell Science.
Jobst KA (1995) A Critical Analysis of Acupuncture in Pulmonary Disease: Efficacy and
Safety of the Acupuncture Needle. J Alt Compl Med 1: (1) 57â€”85.
Jobst KA (1996) Acupuncture in Asthma and Pulmonary Disease: An Analysis of
Efficacy and Safety. J Alt Compl Med 2 (1): 179â€”206.
Jobst KA et al (1986) Controlled trial of acupuncture for disabling breathlessness. Lancet Dec
Joos S et al (2000) Immunomodulatory Effects of Acupuncture in the Treatment of Allergic
Asthma: A Randomized Controlled Study. J Alt Compl Med 6(6): 519-525.
Kleijnen J, ter Riet G, Knipschild P (1991) Acupuncture and asthma: a review
of controlled trials. Thorax 46: 799 - 802.
Lai Xinsheng (1993) Observation of the Curative Effect of Acupuncture on Type I Allergic
Diseases. Journal of Traditional Chinese Medicine 13 (4): 243-248
Landa N M, Fadeeva M A ( 1992) Acupuncture effect on re-activity and some indices of
hormonal systems of children suffering from bronchial asthma, pollinosis, and atopic
dermatitis. British Journal of Acupuncture15 (1): 3-8.
Lewith GT, Watkins AD. (1996) Unconventional therapies in asthma: an
overview. Allergy 51: 761 - 769.
Linde K et al (1996) Randomised Clinical Trials of Acupuncture for Asthma - a Systematic
Review. Forsch Komplementarmed 3 (3): 148 - 155.
Linde K, Jobst K, Panton J (2001) Acupuncture for the treatment of chronic asthma
(Cochrane Review) in The Cochrane Library, Issue 2,2001 Oxford: Update Software Ltd.
Mitchell P, Wells JE (1989) Acupuncture for Chronic Asthma: A controlled trial with six
months follow-up. Am J. Ac. 17 (pt 1): 5-13.
Tashkin DP et al (1977) Comparison of real and simulated acupuncture and isoproterenol in
metacholine-induced asthma. Annals of Allergy. 36 (6) : 379-387.
Tashkin DP et al (1985) A controlled study of real and simulated acupuncture in the
management of chronic asthma. J. Allergy and Clin. Immun. 76 (6): 855 - 864.
Shao Jingming (1985) Clinical Observation on 111 cases of Asthma Treated by Acupuncture
and Moxibustion. Journal of Traditional Chinese Medicine. 5(1): 23-25.
Yu DYC, Lee SP (1976) Effect of Acupuncture on Bronchial Asthma. Clin. Sci. & Mol. Med.
51: 503 - 509
Zang Junqi (1990) Immediate Antiasthmatic Effect of Acupuncture in 192 Cases of Bronchial
Asthma. Journal of Traditional Chinese Medicine. 10(2): 89-93
Grateful acknowledgement is made to Jennifer Dale for her work in preparing this briefing
Find a local acupuncturist
To search by other criteria - name, town - click here
Bronchial asthma and acupuncture: the evidence for effectiveness
Terms and conditions:
Terms and conditions The use of this fact sheet is for the use of British Acupuncture Council members and is subject to the strict conditions imposed by the British Acupuncture Council details of which can be found in the members area of its website www.acupuncture.org.uk.
In the news
World Health Organisation
The World Health Organisation lists a wide variety of diseases or disorders for which acupuncture therapy has been tested in controlled clinical trials