Generated by Codex with GPT-5

This summary covers The Economist’s May 9th, 2026 Science & technology column Well Informed: Does acupuncture work?, published under the headline Does acupuncture work?.

The article gives acupuncture a measured verdict: it looks useful for some kinds of pain, but the evidence becomes much weaker once the claims move beyond pain relief. That distinction matters because acupuncture occupies an unusual place in modern medicine. It is an ancient Chinese practice traditionally explained through the flow of qi, or vital energy, yet it is now marketed by influencers, athletes and wellness clinics for everything from anti-ageing to fertility, asthma, anxiety and muscle recovery.

The Economist’s point is not that all of this is nonsense, nor that acupuncture has been fully validated. It is that the evidence has to be sorted by condition. Some claims rest on large randomized trials and repeatable findings. Others rest on weaker studies, plausible-sounding anecdotes or effects that may come from the ritual of treatment rather than the needles themselves.

The Strongest Case Is Pain

The clearest support comes from chronic pain. A 2018 analysis in the Journal of Pain combined results from 39 randomized trials involving 20,827 patients with shoulder pain, chronic musculoskeletal pain, headaches or osteoarthritis. Patients received either traditional acupuncture, sham acupuncture or no acupuncture. More than four weeks after treatment began, those who had real acupuncture reported less pain than people in the comparison groups. The benefit also appeared to last: it had not faded much a year later.

Later studies have broadly supported that finding. This does not mean acupuncture is a cure, or that every patient will benefit. It does mean the practice has enough evidence behind it to be taken seriously as a pain-management option, especially when compared with drugs that can bring stronger side-effects or dependency risks.

One possible mechanism is physical rather than mystical. Helene Langevin, a former director of the National Centre for Complementary and Integrative Health at America’s National Institutes of Health, has studied whether needles twist fascia, the connective tissue that runs through the body. That tugging could stimulate nerve endings in ways that reduce pain. The hypothesis does not prove acupuncture’s traditional theory, but it offers a biologically plausible explanation for at least part of its effect.

The Placebo Problem

The harder question is how much of acupuncture’s benefit comes from the procedure itself and how much comes from the patient’s expectation that something meaningful is being done. Placebo effects are not imaginary in the simple sense; the brain can genuinely dampen pain when it believes an intervention is real. More elaborate interventions often produce stronger placebo responses, and acupuncture is certainly elaborate. It involves expertise, touch, needles and a ritual that feels more substantial than swallowing a sugar pill.

That makes the trial evidence difficult to interpret. A 2020 paper in JAMA Internal Medicine found no significant difference in pain relief between true acupuncture and sham acupuncture. If shallow or misplaced needles can produce similar results, the specific location and traditional theory of the needles may matter less than the clinical encounter around them.

This does not make acupuncture useless. For a patient in pain, relief is relief, provided the treatment is safe and not replacing necessary care. But it does complicate claims that acupuncture works because ancient theory has identified special points on the body. The more cautious conclusion is that acupuncture may help some pain patients through a mix of physiological stimulation, expectation and therapeutic ritual.

Claims Beyond Pain Need More Proof

Outside pain, the evidence thins. A 2022 review in Complementary Therapies in Medicine looked across 862 systematic reviews and meta-analyses. It reported that acupuncture may reduce post-operative nausea about as well as some anti-nausea drugs. It also found possible benefits for migraines and tension headaches, cancer-related fatigue, female infertility when paired with medical reproductive treatment and chronic pelvic pain in men.

Even that relatively positive review came with reasons for caution: it was written by practising acupuncturists and funded by the International Society of Chinese Medicine. More important, the broader evidence base was uneven. For 86 other conditions, including claims related to muscle recovery, the trials were not strong enough to demonstrate a positive effect. For another six ailments, no effect was found.

The article’s final judgment is therefore practical rather than ideological. Acupuncture is a reasonable intervention for chronic pain, partly because it appears safer than many pharmaceutical alternatives. But the wider wellness industry has run far ahead of the evidence. Patients may reasonably consider acupuncture for pain, migraines, nausea or a few other areas where research is more promising. They should be far more skeptical when it is sold as a general-purpose remedy.

The takeaway is that acupuncture should be judged like any other treatment: by condition, by evidence and by risk. Its age does not prove that it works. Its association with traditional medicine does not prove that it cannot. The strongest case is narrow but real. The rest remains mostly unproven.