How does acupuncture work? The best-supported answer is through the nervous system. Fine needles stimulate sensory nerves in skin and muscle, which damps pain signalling at the spinal cord, prompts an endogenous opioid response and causes small chemical changes in the tissue around each needle. This guide gives UK physiotherapists, sports therapists and clinic teams the straight version: the mechanisms, the trial evidence, what NICE recommends and where the research runs out.
TL;DR
- Acupuncture's measurable effects are best explained by the nervous system: gated signals at the spinal cord, the release of the body's own opioid chemicals, and local tissue changes such as adenosine release.
- The largest pooled analysis of trial data (39 trials, 20,827 patients) found acupuncture beat both sham needling and no-treatment controls for chronic pain, and most of the benefit was still there a year on.
- The margin over sham is modest, so expectation and the ritual of care carry part of the effect. That is worth saying to patients rather than hiding from them.
- NICE recommends considering a course of acupuncture for chronic primary pain and for chronic tension-type headache. It reached that position in a review that rejected TENS and therapeutic ultrasound outright.
- No anatomical structure corresponding to meridians has ever been found, and you do not need one to explain the measured effects.
- For clinics, Meglio sterile acupuncture needles start at £4.16 ex VAT for a box of 100, falling to £3.12 a box (3.12p a needle) when you buy 40 or more boxes.
Why the question needs a straight answer
Ask ten patients on a physio caseload what they make of acupuncture and you get the full spread. Some swear by it, some assume it is nonsense, and most sit in the middle, curious but wary. Practitioners who needle get asked how it works every week, and the two easy answers, "it rebalances your energy" and "it is just placebo", are both wrong.
That leaves clinicians with a communication problem. Overstate the mechanism and you lose the patient the moment they read a sceptical article. Undersell it and you talk them out of a treatment NICE says is worth considering for chronic primary pain. The defensible answer sits in the middle, and it rests on a deeper evidence base than most people expect.
This article is written for the practitioners who needle, and for the clinic owners deciding whether needling belongs on the treatment menu.
How does acupuncture work? The mechanisms with real support
Set the meridian charts aside for a moment and follow the needle. When a fine filiform needle passes through skin into muscle, researchers can measure three separate responses.
Gate control: a competing signal at the spinal cord
The oldest explanation still standing comes from gate control theory, first proposed in the 1960s. Pain signals travelling to the brain pass through a kind of gate at the dorsal horn of the spinal cord, and other sensory input arriving at the same segment can close it. A needle is precisely that kind of input: novel, focused and non-damaging. The same principle explains why rubbing a knocked elbow helps. Gating accounts well for the relief patients feel during and shortly after a session, though on its own it cannot explain why trial benefits persist months after the last needle comes out.
The endogenous opioid response
The second mechanism runs deeper. Needling appears to engage the body's descending pain modulation system. Signals climb from the needle site to the brainstem, which responds by releasing serotonin, noradrenaline and opioid peptides that suppress incoming pain traffic at the spinal cord. This is the mechanism most researchers reach for when explaining why effects outlast the session itself, because repeated exposure can recalibrate how the nervous system processes pain.
Local tissue effects around the needle
The third mechanism is local. A 2010 study in Nature Neuroscience, carried out in mice, measured a sharp rise in adenosine, a compound with pain-damping properties, in the tissue around the needle. The pain relief disappeared in animals lacking adenosine A1 receptors, and slowing the breakdown of adenosine made the relief last longer. Animal findings never transfer to humans automatically, but this is direct mechanistic evidence that something chemical, local and specific happens at the needle site.
What the higher-quality evidence supports
The single most useful reference for a clinician is the Acupuncture Trialists' Collaboration meta-analysis, updated in 2018 in the Journal of Pain. It pooled individual patient data, not just trial summaries, from 39 randomised trials covering 20,827 patients with chronic musculoskeletal pain, osteoarthritis, chronic headache or shoulder pain. Acupuncture was superior to both sham and no-acupuncture controls for every condition tested, with overall effects of around half a standard deviation against no treatment and around 0.2 against sham. The effects held: at one year, roughly 85% of the treatment effect remained. The authors' conclusion is worth quoting to sceptical colleagues, that decreases in pain after acupuncture "cannot be explained solely in terms of placebo effects".
For headache specifically, the Cochrane review of acupuncture for episodic migraine prevention pooled 22 trials with 4,985 participants. Adding acupuncture to symptomatic treatment reduced headache frequency, it held a small but genuine edge over sham, and it performed comparably to preventive drugs while producing fewer adverse effects.
Put together, that is a picture of moderate, durable relief for chronic pain and headache. Not a miracle. Better than nothing and better than sham, by a measurable margin.
What the evidence does not support
Now the other side of the ledger. No dissection, scan or biopsy has located a meridian, and qi is not something instruments can detect. None of the mechanisms above requires either concept, and the trial evidence neither confirms traditional theory nor depends on it.
The sham gap deserves honest handling too. In the pooled data, the gap between acupuncture and sham needling (wrong points, or needles that do not penetrate) was much smaller than the gap between acupuncture and no treatment, which means the sham procedure itself delivered a good share of the measured benefit. Part of what acupuncture provides comes from expectation, attention and the structure of care, which is true of most hands-on treatment. The modest margin over sham also suggests that precise point location matters less than traditional charts imply.
Finally, scope. The strong evidence covers chronic pain and headache. The NHS lists long-term pain, migraine-type headaches and severe hiccups as the conditions where it may offer acupuncture. Claims reaching well beyond pain run ahead of the data, and practitioners who repeat them spend credibility they will want back later.
What NICE and the NHS say
NICE's chronic primary pain guideline, NG193, recommends that clinicians "consider a single course of acupuncture or dry needling, within a traditional Chinese or Western acupuncture system" for people aged 16 and over. The conditions attached are practical rather than clinical: the course should be delivered in a community setting, by a band 7 or lower healthcare professional with appropriate training, and made up of no more than 5 hours of professional time, or delivered by another appropriately trained professional at equivalent or lower cost. The same guideline tells clinicians not to offer TENS, therapeutic ultrasound or interferential therapy for chronic primary pain at all, because there is no evidence of benefit. Acupuncture surviving a review that culled those clinic staples says something.
For headache, CG150 recommends considering a course of up to 10 sessions over 5 to 8 weeks for the prophylactic treatment of chronic tension-type headache. For migraine prevention, the guideline (amended in 2025) places acupuncture after the drug options: where propranolol, topiramate and amitriptyline have not worked, are not tolerated or are unsuitable, it suggests considering the same course of up to 10 sessions over 5 to 8 weeks.
NHS guidance on complementary therapies reflects the same picture. Acupuncture may be offered on the NHS for long-term pain, migraine-type headaches or severe hiccups, and the NHS warns it may not be safe for people taking blood-thinning medicine, at increased risk of bleeding, or with a skin condition or skin infection.
What this means in clinic
For patients, the two-sentence version works: needling gives your nervous system a strong, harmless input that changes how pain signals are processed, and in large trials that produced reductions in chronic pain which lasted well beyond the course. That framing earns trust, and it survives contact with whatever the patient reads next.
On delivery, follow the NICE course structure: up to 10 sessions, with results judged at the end of the course rather than session by session. Point safety belongs in the same conversation. Our guide to the five forbidden acupuncture points covers the sites practitioners avoid and why.
On kit, the variables that matter are gauge, length and coating, matched to the region you treat. Our clinic guide to choosing acupuncture needles by gauge, length and coating walks through the combinations.
Meglio's sterile acupuncture needles come in boxes of 100, with diameters from 0.20mm to 0.40mm and lengths from 13mm to 75mm. A box costs from £4.16 ex VAT, and volume discounts start at eight boxes and reach 25% at forty, which takes that box down to £3.12 ex VAT, or 3.12p a needle. Clinics working through hundreds of needles a week can browse the full acupuncture range or read our wholesale buying guide for the order-size arithmetic.
FAQs
What does acupuncture do?
Acupuncture stimulates sensory nerves in skin and muscle. That input damps pain signalling at the spinal cord, prompts the release of the body's own opioid chemicals and causes small local tissue changes around the needle. In trials, the measurable result is a moderate, durable reduction in chronic pain and in headache frequency. It does not correct energy flow, because no structure carrying such energy has ever been found.
Does acupuncture hurt?
Not much, for most people. Acupuncture needles are far finer than injection needles (Meglio's range from 0.20mm to 0.40mm in diameter), so insertion is often not felt at all. Many patients notice a brief dull ache or heaviness once the needle reaches muscle, which practitioners regard as a normal response. Sharp or lasting pain is not normal, and the needle should be repositioned or removed.
Is acupuncture safe?
In trained hands, yes. Serious adverse events are rare, and the Cochrane migraine review recorded fewer side effects with acupuncture than with preventive drugs. The NHS advises that acupuncture may not be safe for people taking blood-thinning medicine, at increased risk of bleeding, or with a skin condition or skin infection. Single-use sterile needles and sound point selection are the practitioner's side of the bargain.
How many sessions does acupuncture take?
NICE recommends courses of up to 10 sessions over 5 to 8 weeks for chronic tension-type headache, and its chronic primary pain guideline caps a course at 5 hours of clinician time, which works out at a similar number. Some people feel a change after one or two sessions, but the trial benefits come from completed courses, so judge results at the end. What patients do between sessions matters too; see our guide to recovery after acupuncture.
Does acupuncture work for chronic pain?
The best available data says yes, moderately. Pooled individual patient data from 39 trials covering 20,827 patients found acupuncture outperformed both sham needling and no-treatment controls for musculoskeletal pain, osteoarthritis and chronic headache, and about 85% of the benefit remained a year later. The margin over sham is modest, so how acupuncture works should be explained honestly, but the effect is not placebo alone.
What is the difference between acupuncture and dry needling?
The needle is the same; the framework differs. Dry needling targets myofascial trigger points using Western anatomical reasoning, while traditional acupuncture selects points from Chinese medicine theory. NICE treats them together for chronic primary pain, recommending a course of either. Many UK physios train in Western medical acupuncture, which sits between the two. Our dry needling guide covers the technique in detail.
Conclusion
So, how does acupuncture work? Through the nervous system: gated signals at the spinal cord, an opioid response driven from the brainstem, and local chemical changes around the needle. The pooled trial data shows moderate relief for chronic pain and headache that holds up at one year, NICE backs defined courses for both, and none of it requires meridians to be real. Most patients would rather have that answer than either extreme, and it is the one the evidence lets you give.
This article is intended for qualified healthcare professionals and is not a substitute for clinical training or professional judgement. Always apply evidence-based practice and refer patients to appropriate specialists where required.
