Acupuncture treatment for pain: systematic review of randomised clinical trials with acupuncture, placebo acupuncture, and no acupuncture groups
Madsen et al. · BMJ · 2009
Evidence Level
STRONGOBJECTIVE
To assess the analgesic effect of real acupuncture compared with placebo and with no treatment
WHO
3,025 patients with various painful conditions
DURATION
Ranged from 1 day to 12 weeks
POINTS
Traditional points vs. sham points vs. non-penetrating needles
🔬 Study Design
Real Acupuncture
n=1206
needling at traditional points with stimulation
Placebo Acupuncture
n=943
superficial or sham needling
No Acupuncture
n=876
standard care only
📊 Results in numbers
Difference acupuncture vs. placebo
Difference placebo vs. no treatment
Standardized effect acupuncture vs. placebo
Heterogeneity across studies
Percentage highlights
📊 Outcome Comparison
Standardized mean difference in pain
This large review showed that acupuncture has only a small effect on pain relief when compared with placebo. The benefit corresponds to a reduction of only 4 mm on a 100 mm scale, which may not be clinically significant.
Article summary
Plain-language narrative summary
This comprehensive systematic review, published in the BMJ in 2009, represents an important milestone in research on acupuncture for pain. Investigators at the Nordic Cochrane Centre rigorously analyzed 13 randomized clinical trials involving 3,025 patients with diverse painful conditions. The distinguishing feature of this study was the inclusion of only trials with three groups: real acupuncture, placebo acupuncture, and no treatment — a design that allows separation of the specific effects of acupuncture from placebo effects. The conditions studied included knee osteoarthritis, headaches, migraine, low back pain, fibromyalgia, postoperative pain, and procedural pain.
Treatment duration ranged from a single session to 12 weeks of treatment. The methodology was rigorous, with comprehensive searches across multiple databases and careful assessment of the quality of included studies. The results revealed important and controversial findings. When compared with placebo acupuncture, real acupuncture showed only a small statistically significant benefit, with a standardized mean difference of -0.17.
Translated into clinical terms, this represents a reduction of only 4 mm on a 100 mm visual analog scale — a benefit of questionable clinical relevance. By contrast, when placebo acupuncture was compared with no treatment, a moderate effect was found, corresponding to a 10 mm reduction in pain. This finding is particularly intriguing, as it suggests that much of the benefit attributed to acupuncture may be related to the psychological effects of the treatment ritual, the therapist-patient interaction, and the expectation of improvement, rather than to specific effects of needling at traditional points. The analysis also investigated whether different types of placebo (non-penetrating needles vs.
superficial needling at sham points) influenced the results, but found no significant association. Surprisingly, there was a trend toward greater acupuncture effects when the placebo was penetrating, contrary to expectations. The authors conducted several sensitivity analyses that confirmed the main findings, including analyses restricted to studies of high methodological quality and trials with experienced acupuncturists. Heterogeneity across studies was substantial in the placebo vs.
no treatment comparison, indicating that the placebo effect varies substantially across different contexts. The clinical implications are significant. The study suggests that acupuncture may have only minimal specific effects on pain, and that much of the reported benefit may be attributable to placebo effects. This does not entirely invalidate acupuncture as a therapeutic option, but raises important questions about its mechanisms of action and actual efficacy.
The authors acknowledge important limitations, including the impossibility of full blinding of acupuncturists and variability in treatment protocols. They also note that their findings apply to the additive effects of acupuncture beyond standard care, not to acupuncture as a sole treatment.
Strengths
- 1Innovative three-group design allows separation of specific and placebo effects
- 2Large, diverse sample with multiple painful conditions
- 3Rigorous methodology with comprehensive sensitivity analyses
- 4Systematic search across multiple databases
- 5Careful assessment of methodological quality of studies
Limitations
- 1Impossibility of blinding acupuncturists in all studies
- 2Substantial heterogeneity across studies for some outcomes
- 3Variability in acupuncture protocols and types of placebo
- 4Possible reporting bias in no-treatment groups
- 5Limited to additive effects of acupuncture on standard care
Expert Commentary
Dr. Marcus Yu Bin Pai
MD, PhD · Pain Medicine · Physical Medicine and Rehabilitation · Medical Acupuncture
▸ Clinical Relevance
This work from the Nordic Cochrane Centre group has firmly entered the repertoire of any physician who prescribes acupuncture in a pain service. The practical starting point is this: the 4 mm difference on the visual analog scale between real acupuncture and placebo lies below the minimal clinically important threshold, generally accepted at 10-13 mm for chronic pain. The 10 mm difference between the placebo group and the no-treatment group is clinically more robust and directly informs the decision to include acupuncture in multimodal protocols. For populations with knee osteoarthritis, chronic headache, or low back pain — exactly the conditions represented in the review — acupuncture operates as a component of a strategy that mobilizes expectation, therapeutic alliance, and contextual neurobiology, all legitimate, usable mechanisms within a well-structured rehabilitation plan.
▸ Notable Findings
The finding that merits clinical attention is not the small difference between real acupuncture and placebo, but the magnitude of the active placebo effect: 10 mm of pain reduction produced by sham needling, regardless of point or insertion depth. This places the needling ritual — the physical contact, the physician's attention, the expectation of relief — as a therapeutic variable with measurable weight, not noise to be eliminated. Equally noteworthy is the counterintuitive trend toward greater effects of real acupuncture when the placebo was penetrating: this suggests that the conceptual separation between 'specific' and 'non-specific' is more porous than classical models acknowledge. The 36% heterogeneity in the placebo vs. no-treatment comparison further indicates that the clinical context substantially modulates the effect, with direct implications for the design of care protocols.
▸ From My Experience
In my practice in the musculoskeletal pain clinic, what Madsen and colleagues quantified has long been clinically perceptible: patients with moderate chronic pain respond to needling in a way that exceeds any purely mechanical effect on the point. I typically see an initial response — reported pain reduction and functional improvement — from the third or fourth session, with a clinical plateau generally between the eighth and twelfth session for conditions such as chronic low back pain and knee osteoarthritis. I routinely combine acupuncture with supervised exercise and, when necessary, with adjuvant analgesia such as duloxetine or pregabalin, because none of these resources alone is sufficient. I am reserved about patients with unrealistic expectations of cure, or with severe central sensitization syndromes without concurrent pharmacologic management — the response in those cases is frustrating for everyone. What this work reaffirms is that the relational component of treatment is not noise: it is part of the dose.
Full original article
Read the full scientific study
BMJ · 2009
DOI: 10.1136/bmj.a3115
Access original articleScientific Review
Marcus Yu Bin Pai, MD, PhD
CRM-SP: 158074 | RQE: 65523 · 65524 · 655241
PhD in Health Sciences, University of São Paulo. Board-certified in Pain Medicine, Physical Medicine and Rehabilitation, and Medical Acupuncture. Scientific review and curation of every entry in this library.
Learn more about the author →Medical disclaimer: This content is for educational purposes only and does not replace consultation, diagnosis, or treatment by a qualified professional. Some information may be assisted by artificial intelligence and is subject to inaccuracies. Always consult a physician.
Content reviewed by the medical team at CEIMEC — Integrated Centre for Chinese Medicine Studies, a reference in Medical Acupuncture for over 30 years.
Related articles
Based on this article’s categories