Is Acupuncture Pseudoscience? What the Evidence Actually Says

Scope note: This article is educational content for practitioners and interested readers. It is not medical advice, and it does not claim acupuncture treats or cures any condition. Where the evidence is weak or mixed, we say so.

Taking the criticism seriously

Type "acupuncture" into a search engine and it won't take long to find the words pseudoscience, quackery and theatrical placebo. Skeptics make some fair points: the classical concepts of Qi (气) and meridians (经络, jīng luò) predate modern anatomy by two thousand years, some published trials are low quality, and a few practitioners do make claims the evidence can't support.

But "some claims are unproven" is a very different statement from "the whole practice is a placebo." That second claim is testable — and it has been tested, repeatedly, with some of the largest datasets in pain research. Here's what that testing actually found.

The strongest test: 17,922 patients, one at a time

The most rigorous answer to the placebo question comes from the Acupuncture Trialists' Collaboration, an international group that did something unusual: instead of summarising published trial results, they obtained the raw individual patient data from 29 high-quality randomised trials — 17,922 patients in total — and re-analysed it from scratch [1].

Their conclusion, published in the Archives of Internal Medicine (now JAMA Internal Medicine), was direct: acupuncture was superior to both sham (placebo) needling and no-acupuncture controls for back and neck pain, osteoarthritis and chronic headache, with p < 0.001 for every comparison. In the authors' own words, the significant differences between true and sham acupuncture "indicate that acupuncture is more than a placebo" [1].

This matters because of how the analysis was done. Individual patient data meta-analysis is the gold standard of evidence synthesis — it can't be gamed by selective reporting, and only trials with unambiguously adequate allocation concealment were included. A later follow-up confirmed the effects persist over time, with only about a 15% reduction in effect at one year [2].

The animal problem for the placebo theory

Here's a question the "it's all in your head" explanation struggles with: why does acupuncture produce measurable, repeatable effects in animals?

Animals don't read health blogs. A dog with a slipped disc has no expectation of getting better, no cultural belief in needles, and no desire to please its owner. Yet:

• In a randomised trial of dogs with thoracolumbar intervertebral disc disease — a painful, often paralysing spinal condition — electroacupuncture produced significantly better neurological recovery than decompressive surgery alone, and the combination of surgery plus electroacupuncture produced the best outcomes of all [3].
• In horses with chronic back pain, electroacupuncture significantly reduced pain scores compared with a control procedure, measured by blinded assessors using pressure algometry — an objective instrument, not an opinion [4].
• In laboratory animals, the biochemical mechanisms can be measured directly. A landmark study in Nature Neuroscience showed that needling an acupuncture point in mice triggered a local release of adenosine, which acted on A1 receptors to reduce pain — and in mice genetically engineered to lack those receptors, the pain relief simply didn't happen [5]. That's a specific, receptor-mediated biological mechanism, not suggestion.

To be honest and complete: not every animal trial is positive. A well-designed 2020 placebo-controlled crossover trial in dogs with osteoarthritis found no significant difference between acupuncture and placebo on objective force-plate measures [6], and a 2024 evidence review rated the overall canine chronic-pain evidence as weak [7]. The honest summary is that animal evidence is supportive but mixed — strong for neurological conditions like disc disease, weaker for chronic joint pain. What it does rule out is the claim that every positive result in every species is expectation effects.

The mechanism research

Placebo is a real and interesting phenomenon — but it doesn't release adenosine, it doesn't change nerve conduction, and it doesn't show dose-dependence. Acupuncture research has identified several measurable physiological pathways:

• Local purinergic signalling: the adenosine A1 receptor pathway described above [5].
• Endogenous opioid release: electroacupuncture at different frequencies selectively releases different opioids (endorphins, enkephalins, dynorphins) in the central nervous system, an effect blocked by opioid antagonists like naloxone — one of the most replicated findings in the field [8].
• Nervous system modulation: neuroimaging studies show needling produces reproducible changes in limbic and sensorimotor brain regions, distinct from the patterns produced by sham needling [9].

None of this means we understand everything about how acupuncture works. It means the "no plausible mechanism" argument is out of date.

Where the skeptics have a point

Intellectual honesty cuts both ways, and practitioners do their patients no favours by ignoring the legitimate criticisms:

1. Effect sizes against sham are modest. In the Vickers analysis, real acupuncture beat sham by roughly 0.15–0.23 standard deviations [1]. The bigger effects were against no-treatment controls. Context, ritual and the therapeutic relationship are part of how the treatment works — as they are in much of medicine.

2. Blinding is genuinely hard. The person needling always knows whether the needle is real. Sham needles that touch the skin may not be physiologically inert, which can shrink the measured difference.

3. Evidence quality varies enormously by condition. The evidence base is strongest for chronic pain conditions and nausea, and much thinner elsewhere. Claims beyond the evidence should be called out — including when they come from inside the profession.

4. Traditional theory is not anatomy. Qi and meridians are a classical framework for organising clinical observation, not structures you'll find on a dissection table. Modern practitioners can respect the framework's clinical logic without pretending it's a microscope slide.

The honest bottom line

"Pseudoscience" means a practice that refuses testing or survives only by ignoring unfavourable results. Acupuncture has been put through placebo-controlled trials involving tens of thousands of patients, individual-patient-data meta-analysis, animal studies immune to expectation effects, and laboratory mechanism research. The result is a nuanced picture: real, specific, measurable effects that go beyond placebo for several chronic pain conditions [1], plausible and demonstrated biological mechanisms [5][8], and — like most of medicine — plenty of areas where the evidence is still catching up.

That's not quackery. It's what an evolving evidence base actually looks like.

If you're experiencing pain or a health concern, see a registered health practitioner. In Australia, registered acupuncturists are regulated by AHPRA under the Chinese Medicine Board of Australia.

References

1. Vickers AJ, Cronin AM, Maschino AC, et al. Acupuncture for chronic pain: individual patient data meta-analysis. Arch Intern Med. 2012;172(19):1444-1453. PubMed

2. MacPherson H, Vertosick EA, Foster NE, et al. The persistence of the effects of acupuncture after a course of treatment: a meta-analysis of patients with chronic pain. Pain. 2017;158(5):784-793. DOI

3. Joaquim JGF, Luna SPL, Brondani JT, et al. Comparison of decompressive surgery, electroacupuncture, and decompressive surgery followed by electroacupuncture for the treatment of dogs with intervertebral disk disease with long-standing severe neurologic deficits. J Am Vet Med Assoc. 2010;236(11):1225-1230. DOI

4. Xie H, Colahan P, Ott EA. Evaluation of electroacupuncture treatment of horses with signs of chronic thoracolumbar pain. J Am Vet Med Assoc. 2005;227(2):281-286. DOI

5. Goldman N, Chen M, Fujita T, et al. Adenosine A1 receptors mediate local anti-nociceptive effects of acupuncture. Nat Neurosci. 2010;13(7):883-888. DOI

6. Baker-Meuten A, Wendland T, Shamir SK, Hess AM, Duerr FM. Evaluation of acupuncture for the treatment of pain associated with naturally-occurring osteoarthritis in dogs: a prospective, randomized, placebo-controlled, blinded clinical trial. BMC Vet Res. 2020;16:357. Full text

7. Gong J, Thompson MF. Assessing the efficacy of acupuncture as the sole analgesic for canine chronic pain. Veterinary Evidence. 2024;9(1). Full text

8. Han JS. Acupuncture and endorphins. Neurosci Lett. 2004;361(1-3):258-261. DOI

9. Napadow V, Kettner N, Liu J, et al. Hypothalamus and amygdala response to acupuncture stimuli in carpal tunnel syndrome. Pain. 2007;130(3):254-266. DOI