Article Text


Summaries of recent papers
  1. Adrian White1,
  2. Michael Meinen2,
  3. Amer Sheikh3
  1. 1Institute of Translational and Stratified Medicine, Plymouth University Peninsula Schools of Medicine and Dentistry, Plymouth, UK
  2. 2Denton Park Health Centre, Newcastle upon Tyne, UK
  3. 3Ashford Health Centre, Ashford Hospital, Ashford, Middlesex, UK
  1. Correspondence to Dr Adrian White, Department of Primary Care, Peninsula Medical School, N21 ITTC Building, Tamar Science Park, Plymouth PL6 8BX, UK; Adrian.white{at}

Statistics from

Clinical effectiveness trials

Effect of de qi in Bell’s palsy

▸ Xu SB, Huang B, Zhang CY, et al. Effectiveness of strengthened stimulation during acupuncture for the treatment of Bell palsy: a randomized controlled trial. CMAJ 2013;185:473–9.

Randomised controlled trial (n=338) testing the effect of eliciting de qi.


Three hundred and thirty-eight patients with Bell's palsy diagnosed within 1 week of onset of symptoms were randomised to two groups. The acupuncture group had vigorous needle stimulation to achieve de qi while the control group had needle insertion without manipulation in the same predetermined acupuncture points. Both groups received a course of prednisolone in a tapering dose for 2 weeks as standard medical treatment.

Acupuncture points used were GB14, ST4, ST6, ST7, TE17 and LI4 on the affected side. Both groups received 20 sessions lasting 30 min, five sessions per week for 4 weeks.

The primary outcome measure was the House–Brackmann scale, which is a 6-grade objective score for Bell’s palsy (6 is complete paralysis). The scorers were blinded to group allocation. Secondary outcomes were the Facial Disability Index and the World Health Organization’s Quality of Life questionnaire.


At 6-month follow-up, patients in the de qi group had better facial function (OR 4.16), better disability assessment and better quality of life than controls. Complete recovery occurred in 94.3% of patients in the de qi group compared with 77.1% of controls (figure 1), which is a significant difference. The size of the effect was correlated with the strength of the de qi sensation. The control group had the same recovery rate as would be expected with patients who took prednisolone alone.

Figure 1

Numbers making full recovery from Bell's palsy. Based on CMAJ 2013;185:473–9.


At last this is a study that addresses the question whether de qi is necessary for acupuncture to have an effect. The answer from this randomised controlled trial, which has a sound research methodology, is a clear ‘yes’ in the case of Bell's palsy.

The authors estimated that they would need nearly 500 participants to show the effect of de qi but, in practice, the effect was large enough to be seen with only 338.

Specific effect of acupuncture

▸ Karner M, Brazkiewicz F, Remppis A, et al. Objectifying specific and nonspecific effects of acupuncture: a double-blinded randomised trial in osteoarthritis of the knee. Evid Based Complement Alternat Med 2013;2013:427265.

Randomised controlled trial (n=116) comparing single treatments with three different styles of acupuncture.


Patients aged 35–82 with osteoarthritis of the knee were enrolled in three study centres. Participants were given three needling treatments at weekly intervals, one each of individualised classical; ‘modern’ semi-standardised; and ‘non-specific’, in random order. One physician marked up points for all three styles using different coloured pens, then a second (blinded, acupuncture novice) physician needled the points marked by one colour chosen at random. Up to 12 needles were permitted, the same for each style. Points for the ‘modern’ style were ST36, ST34, EX32 twice, SP9, SP10, SP6, GB34, LI4 with optional ah shi, LI3 and ST40. Points for ‘non-specific’ needling were defined non-points elsewhere in the leg. Blinded outcome was assessed by knee flexibility (goniometer) immediately and after 7 days as an objective test and the (subjective) Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score 3 and 7 days after treatment.


Improvement in knee flexibility was significantly higher immediately after classical Chinese acupuncture than in the other groups (figure 2). All methods achieved pain relief, with a patient response rate of 48% for non-specific needling, 64% for modern acupuncture and 73% for classical acupuncture. The data suggest a specific effect of acupuncture in knee mobility and both non-specific and specific effects of needling in pain relief.

Figure 2

Knee flexion. Based on Evid Based Complement Alternat Med 2013;2013:427265.


This trial was a variation of other methods that enable double blinding, although it introduced one true innovation—the use of three coloured pens.

The authors claim it demonstrates that acupuncture has specific effects. This could be true, but we must be objective. Goniometer measurements are notoriously unreliable, which seems not to have been taken into account. Maintaining blinding could have been a problem: classical acupuncture would presumably have involved treatment to several points bilaterally that would appear unrelated to the knee, so both the patient and the second acupuncturist would be at risk of unblinding. Blinding of the patients and practitioners should have been verified.

This could be a comparison of different ‘mechanisms’ rather than different ‘styles’ of acupuncture. We are not told which classical points were used, but they might induce diffuse noxious inhibitory control (DNIC) with short-term effects that are not clinically useful. After all, the pain scores after 1 week showed no differences.

The results are interesting in another way: the differences in knee flexion (goniometer) are more marked than those on WOMAC, suggesting that the subjective changes are an indication of non-specific effects and the objective changes a better indication of specific effects.


▸ Molassiotis A, Bardy J, Finnegan-John J, et al. A randomized, controlled trial of acupuncture self-needling as maintenance therapy for cancer-related fatigue after therapist-delivered acupuncture. Ann Oncol 2013 Published Online First: 21 February 2013. doi: 10.1093/annonc/mdt034

Maintenance acupuncture is advocated by clinicians after successful clinic-based acupuncture. We aimed to assess the effectiveness of maintenance acupuncture in the management of cancer-related fatigue (CRF); treatment delivered by therapists or self-acupuncture/self-needling was compared with no maintenance treatment.


Patients with breast cancer who participated in a randomised trial of acupuncture for CRF management (reported on p 108 of December 2012 issue of Acupunct Med) were re-randomised to receive an additional four acupuncturist-delivered weekly sessions; four self-administered weekly acupuncture sessions (self-needling); or no acupuncture. The primary outcome was general fatigue (Multidimensional Fatigue Inventory). Mood, quality of life and safety were also assessed.


Figure 3 shows the results for these three groups throughout the trial including the initial 6 weeks of acupuncture. 197 patients were re-randomised, with 65 to therapist-delivered sessions, 67 to self-acupuncture/self-needling and 65 to no further acupuncture. Primary outcome scores were equivalent between the therapist-delivered acupuncture and self-acupuncture (p>0.05). A non-significant trend in improving fatigue was observed at the end of 4 weeks in the combined acupuncture arms (p=0.07) compared with no acupuncture. There was no impact on mood or quality of life of the further acupuncture sessions at 18 weeks beyond the improvement observed in the initial trial.

Figure 3

General fatigue scores. Based on Ann Oncol Published Online First: 21 February 2013. doi:10.1093/annonc/mdt034


Self-acupuncture is an acceptable, feasible and safe maintenance treatment for patients with CRF. The study was sized to show an effect equivalent to half an SD (effect size 0.5) and cannot exclude a smaller sustained effect after an initial clinic-based course of acupuncture—which is actually suggested by the trend. The study demonstrated that the effects of the initial course of acupuncture were sustained anyway. Longer follow-up would be more relevant clinically.

Allergic rhinitis

▸ Brinkhaus B, Ortiz M, Witt CM, et al. Acupuncture in patients with seasonal allergic rhinitis: a randomized trial. Ann Intern Med 2013;158:225–34.

Randomised three-arm trial (n=422) comparing acupuncture with sham and medication.


Patients with confirmed seasonal allergic rhinitis (SAR) were randomised to receive acupuncture plus rescue medication (RM), sham acupuncture plus RM, or RM alone. Acupuncture treatment (real and sham) consisted of 12 sessions lasting 20–30 min over a period of 8 weeks. RM consisted of up to two doses of cetirizine 10 mg daily and, if symptoms were not adequately controlled, oral corticosteroids could be used in addition. The patients in the RM alone group were given real acupuncture after the initial 8-week treatment period.

An average of 16 known acupuncture points were used on an individual basis according to TCM theory in the real acupuncture group and stimulated to achieve de qi. Obligatory points included LI4, LI11, LI20 bilaterally and EX-HN3; at least three of EX-HN8, GB20, LR3, LU7, ST36, SP6, TE17 and BL13; and at least three additional points. The sham acupuncture group had needles inserted superficially in five of seven predefined non-acupuncture points bilaterally. Needles were not manually stimulated and de qi was avoided.

Primary outcomes were changes in symptoms and need for medication from baseline. Symptoms were measured using the Rhinitis Quality of Life Questionnaire (RQLQ). Medication use was graded according to the amount taken which they called an RM score. Outcomes were measured at 7–8 weeks after onset of treatment, at 16 weeks and after 1 year.


RQLQ decreased significantly more in the acupuncture group than in the two other groups at the end of the 7–8-week period (figure 4). This was the effect of acupuncture not medication, since the RM score also decreased more in the acupuncture group than in the other two groups.

Figure 4

Rhinitis symptoms. Based on Ann Intern Med 2013;158:225–34.

Between-group differences were smaller and no longer statistically significant at 16 weeks. The acupuncture group showed significantly better scores than the sham acupuncture group in the following year, but not compared with the RM group which had also received real acupuncture.


It is notable that 80% of the participants were self-selected on the basis of response to newspaper articles advertising acupuncture or complementary and alternative medicine for the treatment of SAR. In this study acupuncture was shown to be better than sham or RM during the 8-week treatment period. Longer-term effects are confused by the fact that the RM group was also given real acupuncture after this initial period. It is also interesting to note that the dose of cetirizine (up to 20 mg daily) allowed to be used as RM was higher than the standard recommended dose. Both cetirizine and corticosteroids are potent inhibitors of the allergic response. The fact that acupuncture gave better symptomatic relief over and above the effect of these medications suggests that acupuncture is a viable therapeutic modality in the management of SAR.

Musculoskeletal symptoms from aromatase inhibitor

▸ Bao T, Cai L, Giles JT, et al. A dual-center randomized controlled double blind trial assessing the effect of acupuncture in reducing musculoskeletal symptoms in breast cancer patients taking aromatase inhibitors. Breast Cancer Res Treat 2013;138:167–74.

Randomised controlled trial (n=47) using non-penetrating sham.


Postmenopausal women with early stage breast cancer and experiencing aromatase inhibitor musculoskeletal symptoms (AIMSS) were randomised to eight weekly sessions. Real acupuncture was given at CV4, CV6, CV12 and bilateral LI4, SP6, GB34, ST36, KI3 and BL65, based on a traditional approach. The sham group had non-penetrating needles placed at non-points. The main outcomes were the Health Assessment Questionnaire Disability Index (HAQ-DI) and pain visual analogue scale (VAS). Serum oestradiol, β-endorphin and proinflammatory cytokine concentrations were measured before and after the intervention.


Fifty-one women were enrolled of whom 47 were evaluable; 23 were randomised to real acupuncture and 24 to sham acupuncture. There were significant baseline differences between the groups in HAQ-DI scores (p=0.047), but there was no significant difference in AIMSS changes between real and sham acupuncture (figure 5) or VAS (p=0.31). At week 8, 14/24 (58%) patients in the sham acupuncture group and 8/23 (35%) patients in the real acupuncture group reached the minimal clinically important change in HAQ-DI with a reduction of more than 0.22 units—not significantly different between the groups.

Figure 5

Pain scores. Based on Breast Cancer Res Treat 2013;138:167–74.

Following eight weekly treatments there was a statistically significant reduction in interleukin 17 (p≤0.009) in both groups. No significant modulation was seen in oestradiol, β-endorphin or other proinflammatory cytokine concentrations in either group.


This study involves an important topic: up to 50% of women receiving an aromatase inhibitor complain of AIMSS and 15% discontinue treatment. This trial was stopped for ‘futility’ after 3 years when only 51 of the intended 100 participants had been recruited because an interim analysis showed low power. There were improvements in both groups. Since the improvement might arise from sensory stimulation (whether real or sham needle) rather than point-specific effects of needling, future studies need a third non-acupuncture arm.

Chemotherapy fatigue

▸ Deng G, Chan Y, Sjoberg D, et al. Acupuncture for the treatment of post-chemotherapy chronic fatigue: a randomized, blinded, sham-controlled trial. Support Care Cancer 2013;21:1735–41.


Cancer patients reporting significant fatigue persisting for at least 2 months following the completion of chemotherapy were randomised to receive once weekly true or sham acupuncture for 6 weeks. True acupuncture involved 14 needles to CV4, CV6 and bilateral KI3, ST36, SP6, LI11, HT6 and an auricular point. Non-penetrating sham needles were applied to locations a few millimetres from the same points. Fatigue was evaluated before and after treatment using the Brief Fatigue Inventory (BFI, the primary endpoint). Secondary endpoints included the Hospital Anxiety and Depression Scale (HADS) and Functional Assessment of Cancer Treatment-General (FACT-G) scores.


A total of 1006 patients were assessed and 101 were randomised with 74 evaluated for the primary endpoint (34 true acupuncture, 40 sham controls). BFI scores fell in both groups, but there was no statistically significant difference between the groups (figure 6). HADS and FACT-G scores also improved in both groups, again with no significant difference between groups. Patients in the sham acupuncture group crossed over to receive true acupuncture in week 7. No long-term reduction in fatigue scores was observed at the 6-month evaluation.

Figure 6

Brief Fatigue Inventory (BFI) scores. Based on Support Care Cancer 2013;21:1735–41.


This trial comes hard on the heels of the study by Molassiotis et al1 which showed that acupuncture generates an effect on fatigue that is clinically useful. Deng et al tested the mechanism—how much of the effect is needle-specific and how much is non-specific.

This trial is too small to be really definitive. Sham acupuncture stimulates nerves, so the difference from acupuncture is likely to be small; sample sizes should be about 800. The sample size was calculated to show a clinically relevant difference between groups (1.25 points in BFI), making 44 in each group. In contrast, the study by Molassiotis et al required a sample of 80 in each group. Deng's study also suffered 26% dropouts, so the analysis could be regarded as significantly underpowered: no less than 30% of the acupuncture group (15/49) could not have any change recorded—missing data were replaced (and one case was mistakenly given sham).

The effect seen in the acupuncture arms failed to reach clinical significance with BFI whereas the study by Molassiotis et al reported a relevant effect using the Multidimensional Fatigue Inventory.

Laser to acupuncture points for depression

▸ Quah-Smith I, Smith C, Crawford JD, et al. Laser acupuncture for depression: a randomised double blind controlled trial using low intensity laser intervention. J Affect Disord 2013;148:179–87.

Double-blind randomised controlled trial (n=44).


A randomised double-blind placebo-controlled trial was conducted in Sydney, Australia. Participants aged 18–50 years with DSM-IV major depressive disorder were eligible to join the study. Forty-seven participants were randomised to receive laser acupuncture or placebo laser at LR14, CV14, LR8, HT7 and KI3. The intervention was administered twice a week for 4 weeks and once a week for another 4 weeks, for a total of 12 sessions. The primary outcome was the Hamilton-Depression Rating Scale (HAM-D). The secondary outcomes assessed the change in severity of depression using the Quick Inventory for Depression-Self Reporting (QID-SR), the Quick Inventory for Depression-Clinician (QIDS-CL), with outcomes assessed at 8 weeks. The treatment response (>50% improvement in HAM-D) and remission (HAM-D<8) were analysed.


Improvement on HAM-D was significantly greater in the active laser group on the primary outcome measure (figure 7). Response rates were 18/22 (active) and 4/22 (placebo laser). Results in the secondary measures were similar but were not significantly different between groups. Transient fatigue was the only adverse effect reported.

Figure 7

Hamilton depression score. Based on J Affect Disord 2013;148:179–87.


There is a strong interest in laser acupuncture in Australia, with surprisingly positive results. The beauty of research is the possibility of double blinding. However, some caution is necessary: this is a small trial with no long-term follow-up. This study shows no evidence of mechanisms, but other studies by the author show that laser stimulation of the acupuncture points used activates frontal-limbic-striatal brain regions. Interestingly, the pattern of neural activity is somewhat different for each acupuncture point.

Systematic reviews

Cancer symptoms

▸ Garcia MK, McQuade J, Haddad R, et al. Systematic review of acupuncture in cancer care: a synthesis of the evidence. J Clin Oncol 2013;31:952–60.

Systematic review of clinical trials (n=41) of acupuncture for symptom control.


Searches retrieved 41 randomised controlled trials addressing eight different symptoms. Only one trial was found for each of four additional symptoms (dyspnoea, peripheral neuropathy, hiccups and lower urinary tract symptoms). The retrieved trials were assessed for risk of bias based on factors such as sample size and statistical power, randomisation, description of treatment or problems with blinding participants.


The authors assigned a low risk of bias to only one trial, an unclear risk to seven trials and the rest were judged as high risk, as shown in the table. The evidence was strongest for nausea and vomiting, the only low-bias trial supporting acupuncture, although only five out of the 11 trials were positive. For pain relief, nine of 11 trials were positive, but all scored high or were unclear for potential bias. Five of seven trials addressing hot flushes were reported as positive. For the other symptoms (postoperative ileus, xerostomia, fatigue, sleep disturbance, anxiety/depression) the results were mixed, and all studies were seen as being of lesser quality. The authors conclude that the evidence supports the use of acupuncture only for relief of nausea.

Table Risk of bias assessments


There have been systematic reviews which looked at only one symptom at a time. For a GP or an oncologist, this paper is of value because it gives an overview of a number of symptoms that cancer patients may experience and which may or may not respond to acupuncture. A helpful table is included which explains the basis for the authors’ quality assessment for each included study. The main problem was blinding of participants. A lot has been said and written about this already and I am sure this discussion will continue.


▸ Lee WB, Woo SH, Min BI, et al. Acupuncture for gouty arthritis: a concise report of a systematic and meta-analysis approach. Rheumatology (Oxford) Published Online First: 18 Feb 2013. doi: 10.1093/rheumatology/ket013


All randomised controlled trials involving acupuncture in combination with conventional therapy for gouty arthritis were included. All were assessed for quality using the risk of bias assessment.


Ten randomised controlled trials involving 852 patients with gouty arthritis were included. Meta-analysis (figure 8) showed a significant decrease in uric acid in the treatment group compared with controls. Four studies of 380 patients reported a significant decrease in pain visual analogue scale score in the treatment group.

Figure 8

Acupuncture effect on uric acid. Based on Rheumatology (Oxford) 2013.


This report is worrying. The results of the quality assessment were reported alongside each study in the table of results and not a single study met any of the quality criteria. However, ‘poor quality’ was not mentioned in the Results section, had only five lines in the Discussion section and was not directly mentioned in the conclusion or the Abstract. This reviewer is as attracted to these positive findings for acupuncture as anyone involved in healthcare is bound to be, but there have to be serious doubts about their rigour. The original studies are mostly in Chinese so the amount of checking is limited, but it does look very odd that five of the eight studies in the meta-analysis have exactly the same number of subjects (60), all divided exactly equally (30 in the acupuncture arm and 30 in the medication arm).

Ankle sprain

▸ Park J, Hahn S, Park J-Y, et al. Acupuncture for ankle sprain: systematic review and meta-analysis. BMC Complement Altern Med 2013;13:55.


Randomised controlled trials of acupuncture were included if they involved patients with ankle sprains and reported outcomes of symptom improvement including pain. A Cochrane risk of bias assessment tool was used. Risk ratio (RR) or mean difference was calculated with 95% CIs in a random effects model. Subgroup analyses were performed based on acupuncture type, grade of sprain and control type. Sensitivity analyses were also performed with respect to risk of bias, sample size and outcomes reported.


Seventeen trials involving 1820 participants were included, all published in Chinese. Trial quality was generally poor, with just three reporting adequate methods of randomisation and only one a method of allocation concealment. Significantly more participants in acupuncture groups reported global symptom improvement compared with no acupuncture control groups, as shown in figure 9. Acupuncture as an add-on treatment also improved global symptoms compared with other treatments only, without significant variability (RR 0.61, 95% CI 0.51 to 0.73, I2=1%). Most studies were biased, but the effect of acupuncture remained significant when the analysis was limited to two studies with a low risk of bias.

Figure 9

Global assessment of ankle sprain. Based on BMC Complement Altern Med 2013;13:55.


At least there were some good quality studies which were positive, so it is fair to say the evidence is promising and justifies further studies.

Basic research

Meta-analysis of MRI studies

▸ Chae Y, Chang DS, Lee SH, et al. Inserting needles into the body: a meta-analysis of brain activity associated with acupuncture needle stimulation. J Pain 2013;14:215–22.


We reviewed MRI studies of the functional brain responses that occur because of needle insertion into the body compared with touch. The hypothesis that there are point-specific effects was ignored. An ‘activation likelihood estimation’ meta-analysis was carried out (figure 10).

Figure 10

Three acupressure stimulations (before c, e and g): fall in heart rate from e onwards. Based on eCAM 2013. doi:10.1155/2013/763631.


A total of 28 functional MRI studies consisting of 51 acupuncture and 10 tactile stimulation experiments were included. Following acupuncture needle stimulation, responses formed a common pattern. Activation occurred in the sensorimotor cortical network, which includes the insula, thalamus, anterior cingulate cortex and primary and secondary somatosensory cortices. Deactivation occurred in the limbic-paralimbic neocortical network, which includes the medial prefrontal cortex, caudate, amygdala, posterior cingulate cortex and parahippocampus. The two networks together have been called the ‘pain matrix’. Following control tactile stimulation, weaker patterns of brain responses were detected in areas similar to those stated above. The activation and deactivation patterns following acupuncture stimulation suggest that the haemodynamic responses in the brain simultaneously reflect the sensory, cognitive and affective dimensions of pain.


A significant innovation in this study was to combine the effects of stimulation to any point, ignoring point specificity. The result was to identify a common pattern. Limitations are that the method makes certain assumptions which may not be correct in all cases. The interpretation of the results could also be limited since the MRI studies reviewed only measured acute effects of acupuncture.

Heart rate

▸ Wang L, Cheng W, Sun Z, et al. Ear acupressure, heart rate, and heart rate variability in patients with insomnia. Evid Based Complement Alternat Med 2013;2013:763631.

Case series (n=31) to observe the effects of ear acupressure on heart rate (HR) and HR variability (HRV) in chronic insomniacs.


Patients diagnosed with chronic insomnia had continuous ECG recording during the course of this basic science experiment.

A Semen vaccariae seed was attached to the Shen Men point of the left ear by an adhesive plaster in each patient. Auricular acupressure was applied over this site for 15 s each time, with application of pressure twice per second, making a total of 30 pressure applications per stimulation. This was performed every 10 min for a total of three times during the measurement period in each patient.

HR and HRV were measured in three stages consisting of 5 min phases: two phases before stimulation, four phases during acupressure stimulation and two phases after acupressure.


There was a significant reduction in HR (p<0.5) which started after the second acupressure stimulation and this lasted until the end of the recording period 20 min later. Minimum HR was reached after the third acupressure stimulation. HR increased again towards the end of the recording period, but this was still significantly less than than in the pre-stimulation period.


This study demonstrated an effect on lowering HR using acupressure stimulation of the ear. Since there was no control group, this study cannot demonstrate that this effect is specific for patients with insomnia nor that it is point-specific. Anatomically, the Shen Men area of the ear is supplied by branches of the trigeminal nerve which has parasympathetic nerve connections and provides a possible mechanism for this observed effect.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.