Acupuncture Therapy Safety

Acupuncture Therapy for Acute Headache and Migraine

Trials published since 2016 include a double-blindplacebo-controlled trial in which intraoperative electricalstimulation of acupuncture points reduced intraoperativeopioid requirements, postoperative pain, and duration ofstay in the post-anesthesia care unit [45]. Another double-blind placebo-controlled trial showed that the addition oftranscutaneous electrical acupoint stimulation (applicationof electrical stimulation at acupuncture points) to usualcare for minimally invasive lung cancer surgery reducedpain, reduced patient-controlled intravenous analgesicattempts, and reduced nausea and vomiting, which sup-ports transcutaneous electrical acupoint stimulation as afeasible option for sedation and postoperative analgesia inthoracoscopic pulmonary resection [46].Additionally, acupuncture was effective, safe, andwell tolerated for post-tonsillectomy pain in children,with no significant side effects [47]

Acupuncture Therapy for Postoperative PainIn multiple SRMs, acupuncture was found to be effectivein reducing postsurgical pain compared with sham acu-puncture, controls, and usual care, with a reduction inopioid need (21% opioid reduction at 8 hours, 23% at24 hours, and 29% at 72 hours post surgery) and with alowered incidence of opioid-related side effects such asnausea, dizziness, sedation, pruritus, and urinary reten-tion [34–36]. An SRM found that acupuncture after totalknee arthroplasty reduced pain and was associated withdelayed opioid use [37]. An SRM of perioperative auricu-lar acupuncture found reduced postoperative pain andneed for analgesic use compared with sham or standard-of-care controls [38]. Another SRM on perioperative au-ricular therapies found benefit for pain and intraopera-tive body mass–adjusted fentanyl amount but did notfind prolonged time to first analgesic request in total hiparthroplasty [39]. An SRM of auricular acupuncture foracute and postsurgical pain found that it provided imme-diate pain relief and benefit at 48 hours, was equivalentto analgesics, and had fewer side effects [40]. These find-ings indicate the potential to reduce hospital readmissiondue to uncontrolled pain. (See Table 1.

Medical pain management is in crisis. Pain is pervasiveand has been inadequately addressed by strategiesincluding the escalation of prescription opioids that cre-ated a tragic increase in overdose deaths, addiction, anddiversion.

=0.04). This is contrary to what one wouldhave expected, and we regard it as a chance finding. Wenote that our meta-regression was based on a subjectiveranking of the possibility of a physiological effect ofplacebo, and that both the subgroup analysis and themeta-regression are observational in nature. However,our findings are similar to that of a randomised trialreporting no difference in analgesic effect betweenthree types of placebo acupuncture: acupunctureconsidered specific for another disease, needle inser-tion at non-acupuncture points, and non-penetrativesimulated acupuncture.18We found a tendency for an increase in the use ofanalgesic drugs in the no acupuncture groups com-pared with the placebo and acupuncture groups, whichwould tend to underestimate the effect of placeboacupuncture. We found no tendency for any differencein use of concomitant treatment between the placebogroups and the acupuncture groups.Our sensitivity analyses of the authors’ primaryoutcomes found slightly larger effects of acupunctureand placebo, as well as more heterogeneous results.However, the trials had very dissimilar primaryoutcomes (such as days with headache and number ofanalgesic doses) and primary outcomes in clinical trialsare often changed retrospectively.14 We excluded onetrial as a clear outlier, but the proportion of excludedpatients was small and had little effect on our effectestimates.Other studiesOur finding of limited, at best, analgesic effects ofacupuncture corresponds with the seven Cochranereviews on acupuncture for various types of pain,which all concluded that no clear evidence existed of ananalgesic effect of acupuncture.19-25 Most stressed themethodological shortcomings of the included trials.Our finding of a moderate difference betweenplacebo acupuncture and no acupuncture (standar-dised mean difference −0.42) agrees fairly well with ourprevious review of the effect of placebo in general.9 10Although we previously found an overall difference instandardised mean difference of −0.25 for pain, we alsosaw a tendency for larger effects when the placebointervention was procedural—for example, a shamacupuncture needle (standardised mean difference−0.33)—and not merely a placebo tablet (standardisedmean difference −0.20).Meaning of our reviewInterpreting a standardised mean difference clinicallymay be challenging. On the basis of the mean standarddeviation from the trials that had used visual analoguescales, the effect of acupuncture (standardised meandifference −0.17, −0.26 to −0.07) corresponds to areduction of 4 (2 to 6) mm on a 100 mm scale. The effectof placebo acupuncture (standardised mean difference−0.42 (−0.60 to −0.23), corresponds to a reduction of 10(6 to 15) mm.Attempts at defining a clinically minimal painimprovement have reached quite different conclusionsand have often reported percentage improvement andnot an absolute effect size as we have.26 27 However, aconsensus report characterised a 10 mm reduction on a100 mm visual analogue scale as representing a“minimal” change or “little change.”27 Thus, theapparent analgesic effect of acupuncture seems to bebelow a clinically relevant pain improvement.Our pooled effect of placebo acupuncture (standar-dised mean difference −0.42) is based on trials witheffects that vary much more than is expected by chance.Some of the large trials report an effect of placeboacupuncture that is of clear clinical relevance—forexample, standardised mean difference −0.95,w2 corre-sponding to 24 mm on a 100 mm visual analogue scale—whereas others find effects that seem to be of limitedRESEARCHpage 6 of 8 BMJ | ONLINE FIRST | bmj.comProtected by copyright, including for uses related to text and data mining, AI training, and similar technologies. .by gueston 26 September 2025https://www.bmj.com/Downloaded from27 January 2009.10.1136/bmj.a3115 onBMJ: first published as

We found a small difference between acupuncture andplacebo acupuncture and a moderate differencebetween placebo acupuncture and no acupuncture.The effect of placebo acupuncture varied considerably.Strengths and weaknessesOur review is the first that identifies and analyses threearmed trials of acupuncture for pain, thus providing anestimate of the general analgesic effect of acupunctureand its direct comparison with the analgesic effect ofplacebo acupuncture. The review is fairly large,includes several trials of high methodological quality,and covers a broad range of common painful condi-tions. Furthermore, our main results were similar tothose found in the subgroups of trials with low risk ofbias, in trials using multiple sessions of experiencedacupuncturists choosing acupuncture points at theirdiscretion, and when we analysed the primary out-comes of the trials (instead of the outcome we hadchosen).

Type of placebo acupunctureWe ranked the various placebo acupuncture inter-ventions on a 1-5 scale, where 1 represents a placebotreatment that was most likely to produce physiologicaleffects. We ranked needling at acupuncture pointswithout electrical stimulation but indicator lights on as1w3 w6 ; needling at non-acupuncture points with elec-trical stimulation as 2 w5; superficial needling at non-acupuncture points (20-50 mm) avoiding Qi andmanual stimulation as 3w1 w2 w7 w8 w10-w12; non-penetrat-ing needle as 4w9; and laser turned off, held over thesymptomatic points without using any mechanicalpressure as 5.w4A meta-regression of the 12 trials found nostatistically significant relation between the type ofplacebo intervention and the effect of acupuncture(P=0.60). Supplementary subgroup analyses found astatistically significant difference in effect of acupunc-ture between the two trials using non-penetrative

The search included 234 trials eligible for our updatedCochrane review (in progress) of all types of placebointerventions.9 10 From this sample we identified 20potentially eligible trials for this review. We excludedseven trials—six because they studied transcutaneouselectrical nerve stimulation and one because theintervention was manual acupressure. We included13 trials of acupuncture for pain (3025 patients)(table 1).w1-w13The clinical conditions were knee osteoarthritis,w7-w9tension type headache,w1 migraine,w2 low back pain,w10-w12 fibromyalgia,w4 abdominal scar pain,w13 postoperativepain,w3 w6 and procedural pain during colonoscopy.w5The duration of treatment varied from one day to12 weeks (table 1).Eight trials had clearly concealed the allocation ofpatients.w1 w2 w7-w11 w13 No trials reported blinding of theclinicians managing the acupuncture and placeboacupuncture treatments, whereas blinding of thepatients was explicitly reported in 10 trials.w1-w3 w6-w12In five trials the acupuncture treatment involvedmultiple sessions with experienced acupuncturistswho could choose additional acupuncture points attheir discretion.w1 w2 w4 w7 w11Table 2 describes the placebo treatments. In two trialsthe placebo procedures consisted of non-penetrativeneedling. w4 w9 In 11 trials the placebo procedurepenetrated the skin: seven trials used superficial needlingat non-acupuncture points with fine needles, avoiding Qiand manual stimulation, and four trials used other formsof penetrative needling.

n a pre-planned sensitivity analysis, we used theauthors’ primary outcome.14 In unplanned sensitivityanalyses, we also studied the impact of the methodo-logical quality of the trials and of the type ofacupuncture

Data analysis

ntext of traditional acupuncture, practitioners were perceived as engaging people in active participation in the process of change in ways that were integrally bound with continuing to build a therapeutic relationship and continuing patient-centred treatment with acupuncture. In this situation of ongoing support and treatment, increased agency may be more likely to be experienced as positive and empowering.

One patient said that the ‘mental clarity’ gained had made them more aware of their everyday functional limitations and another queried the extent to which their early physical improvement had led to a false sense of security, perhaps contributing to his later re-injury.Three patients saw their health worsen during treatment but did not link this to acupuncture.Physical improvement following acupuncturePatients cited a range of physical changes post acupuncture. Most common were reductions in the amount and/or frequency of their pain and descriptions of ‘reduced medication’ or the fact that they were no longer ‘living on painkillers’.

Most patients also described how they actively engaged in cognitive or behavioural changes in their lives outside their acupuncture treatment sessions. In addition to thinking positively, some perceived that a new self-awareness of stress patterns and ways of thinking had led to their re-evaluating their life choices and dealing with stress more effectively.

Although some people were actively seeking help from a variety of sources, the predominant attitude to health care appeared to be one of unquestioningly accepting referrals to hospital specialists and therapists in the hope that the problem could be diagnosed, ‘fixed’, or eased. Their GP’s suggestion that they participated in the acupuncture trial was simply one more referral and accepted because there was ‘nothing to lose’, or, as Ernie said: ‘I didn’t think a lot of it! I was just hoping’.

A purposive sample of 20 patients was drawn from the 80 participants in a randomised trial of the use of traditional (five-element) acupuncture for treating unexplained symptoms: the CACTUS trial.38 The trial participants were adults, who were frequent attenders (eight or more consultations a year) with MUPS; they were were informed about the trial by participating London GPs. Participants were offered up to 12 individualised sessions of five-element acupuncture over a 6-month period, which was timed to meet their personal needs. These sessions were carried out by eight acupuncture practitioners.The sample for the qualitative study was drawn to include male and females of diverse ages with varied socioeconomic, educational, and ethnic backgrounds. It included patients of each of the eight acupuncture practitioners. This purposive sample of 20 participants, each of whom was interviewed twice, was considered to provide sufficient data on a wide range of experiences while remaining practical in terms of study time and resources.

Reviews of this range of interventions have identified some common factors that appear to be associated with successful management.25,26 These include:taking patients’ symptoms seriously;involving patients in finding acceptable and empowering explanations that integrate psychological and biological factors;considering non-pharmacological treatments that engage patients in an active role; andengaging patients in taking medication if it is appropriate.A primary-care-based review identified the following practitioner skills as key:helping the patient to feel understood;broadening the agenda;making links (explanatory models that link physical and psychological problems); andnegotiating treatment.27

Taking the reduction of the post-operative requirement of opiate dosage as an index for analgesia, they have revealed that the alternative mode stimulation reduced the morphine requirement by 53%, whereas a constant low (2 Hz) or constant high (100 Hz) frequency produced only a 32% or 35% decrease, respectively [10]. Similar results were obtained in clinical studies on low back pain [9] and diabetic neuropathic pain [11].

A combination of electrical stimulation of different frequencies may produce different profiles of neuropeptide release.

Prior to the discovery of endogenous opioids, we focused on various candidate neurotransmitters including monoamines, and found that serotonin (5-HT) was most important among classical neurotransmitters for the mediation of acupuncture analgesia

As was predicted, 2/100 Hz stimulation increased the release of both dynorphin and endomorphin, whereas 2+100 Hz stimulation increased the release of only dynorphin, but not endomorphin. In other words, 2+100 stimulation does not seem to carry the pure 2 Hz information [26]. It is thus obvious that a proper combination of different frequencies may produce a maximal release of a cocktail of neuropeptides for better therapeutic effects.

This gives strong support to the hypothesis that endogenously released dynorphin is indeed analgesic rather than hyperalgesic or paralytic (Fig. 2).

Further, acupuncture significantly improved joint pain, stiffness, and fatigue in patients with breast cancer (Crew et al., 2010, Molassiotis et al., 2012). Recent studies showed that radiation-induced dry mouth in cancer patients, both measured by subjective scores and objective salivary flow rates, could be remarkably improved by acupuncture (Meng et al., 2012). All the results suggested acupuncture could be a safe, effective, and durable therapy for cancer treatment-related side effects.

However, others claimed the addition of acupuncture to a course of physiotherapy provided no additional improvement in pain relief, and only small benefits were observed in true acupuncture group and nonpenetrating acupuncture group

However, due to lack of details in the methodology and nonreproducible results, his studies were ignored for almost 40 years.

Another important innovation during this time was the electrical stimulation employed in acupuncture basic research. Sarlandiere employed electrical stimulation to the inserted needles to see the altered treatment effects in 1825 (Lu & Needham, 1980). Moreover, other physicians, such as Trousseau, Pidoux, and Duchenne, reintroduced the electrical stimulation into the treatment of chronic pain (Willer, Roby, & Le Bars, 1984).

In 1979, FDA classified acupuncture needles as class III (investigational) medical device but allows their clinical use by licensed practitioners (Hammerschlag, 2000). At the same year, WHO listed 43 kinds of diseases and conditions that can be cured by acupuncture

In 1972, the National Institutes of Health (NIH) in the United States gave its first grant to acupuncture research (Ulett, Han, & Han, 1998). I

In cases of specific forms of pain, perhaps one of the most common indications for acupuncture cited by American adults is lower back pain, either acute or chronic. I

Central to historical teachings on acupuncture is the theory of de qi, which describes the connection and interplay between the acupuncture needles and the qi energy of the body. Qi is an untranslatable word that in essence signifies the potential to transform from one state to another state and each state's interconnectedness, but is most commonly described as the body's “vital energy”. Qi is obtained from three different sources throughout life: from air (air qi), from food (grain qi), and via inheritance from one's parents (original qi) [1]. After qi is obtained, it is believed to flow throughout the body from deep organs to the superficial skin by means of interconnecting meridians, with acupuncture points serving as major confluences of these meridians

more and more Americans are turning to acupuncture as a form of medical treatment when conventional treatments fail,

However, this changed rapidly when President Nixon opened the door to Chinese medical practices with his visit to China in 1972. Since then, interest in complementary forms of therapy has escalated as more and more Americans are diagnosed with chronic illness.

acupuncture is a traditional form of healing used for over 2,500 years in China, its widespread use is a relatively new phenomenon in the United States and other western countries