Does Acupuncture Help Chronic Pain? What the Research Shows

Does acupuncture help chronic pain? The research suggests it does, modestly: in the largest pooled analysis, covering 39 trials and 20,827 people, acupuncture reduced pain more than both sham acupuncture and usual care without acupuncture, though its edge over sham was small. This review explains what those studies measured, what they found, and why guidelines still disagree. For background on the practice itself, see the acupuncture modality overview.

What the studies looked at

Most of what is known comes from the Acupuncture Trialists' Collaboration, an international group that gathered the raw data on every participant from high-quality randomized trials, rather than relying on each trial's published averages. This approach, called an individual patient data meta-analysis, lets researchers analyze results consistently across trials. The group included only trials in which allocation concealment, the step that stops researchers from influencing who gets which treatment, was unambiguously adequate.

Their first analysis, published in 2012 in Archives of Internal Medicine, used data from 29 trials with 17,922 people. It covered four conditions: back and neck pain, osteoarthritis, chronic headache and shoulder pain. The 2018 update in the Journal of Pain added 13 trials published up to the end of 2015, reaching 39 trials and 20,827 people across nonspecific musculoskeletal pain, osteoarthritis, chronic headache and shoulder pain. Both compared acupuncture needling with sham acupuncture and with no acupuncture, such as waiting lists or usual care, and measured pain and function.

Two follow-on papers used the same data set. A 2017 analysis in Pain examined 20 trials, with 6,376 people, that tracked outcomes after treatment ended, to see how long effects lasted. A 2019 analysis in the Clinical Journal of Pain tested whether age, sex, pain duration, starting pain severity or psychological distress changed how much people benefited.

What they found

In the 2012 analysis, acupuncture was better than both sham and no acupuncture for every condition. Compared with sham, pain scores were lower by 0.23 standard deviations for back and neck pain, 0.16 for osteoarthritis and 0.15 for chronic headache. Compared with no acupuncture, the differences were larger: 0.55, 0.57 and 0.42 standard deviations. By common statistical convention, around 0.2 is a small effect and around 0.5 a moderate one. These results held up in robustness checks, including checks for publication bias.

The 2018 update reached the same conclusion with more data. Acupuncture was superior to sham and to no acupuncture for each pain condition, with differences close to 0.5 standard deviations against no acupuncture and close to 0.2 against sham. The authors concluded that acupuncture is effective for chronic pain and that its effects cannot be explained solely by placebo, while acknowledging that factors beyond needling the correct points contribute to the overall benefit.

On durability, the 2018 update estimated only about a 15 percent fall in the treatment effect one year after treatment. The 2017 analysis found that about 90 percent of the benefit relative to no-acupuncture controls was still present at 12 months, while the smaller benefit relative to sham fell by about half over the same period.

On who benefits, the 2019 analysis found that only starting pain severity appeared to change the effect, with people in more pain gaining more. Psychological distress had a small influence, results by sex were inconsistent, and age and pain duration showed no strong link. The authors said the data did not justify sorting people into groups who should or should not receive acupuncture.

How strong the evidence is

This is among the stronger evidence bases for any complementary approach. The data set is large, the researchers used raw individual data rather than summary figures, and the collaboration admitted only trials with clearly adequate allocation concealment. Effects were consistent in direction across all four conditions, and the 2012 results held up in checks for publication bias.

The specific effect, the part that sham acupuncture does not reproduce, is small. Much of the overall benefit is shared with sham treatment, which makes it hard to separate needling at traditional points from the attention, expectation and setting of care. The 2018 authors found that differences between trials were driven mainly by what the control group received, not by how acupuncture was delivered. Sham trials using a penetrating needle, and trials with an intensive control treatment, showed smaller effects.

Limits and open questions

The trials covered four conditions, so the findings do not automatically extend to other chronic pain, such as nerve pain or widespread pain. Longer-term data came from only 20 of the trials, and the authors called for more research on long-term outcomes. Blinding is difficult in acupuncture research, because the practitioner always knows which treatment is given and some sham methods still involve needling or skin contact.

Guidelines read the same evidence differently. NICE guideline NG193, published April 7, 2021, says to consider a single course of acupuncture or dry needling for people aged 16 and over with chronic primary pain, but only within limits, such as delivery in a community setting with no more than five hours of clinician time. NICE guideline NG59 on low back pain and sciatica, from 2016, says not to offer acupuncture for that condition. The research tells you that acupuncture beats sham on average; it does not settle whether that edge is worth the time and cost for a given person.

What this means in practice

If you live with chronic musculoskeletal pain, osteoarthritis or frequent headaches, acupuncture is a reasonable option to discuss with your doctor as part of a wider plan, expecting a modest, often lasting improvement rather than a dramatic change. It works alongside, not instead of, the care you already have.

A course is a better test than one session, since the trials studied courses of treatment. If you are weighing options, the acupuncture or massage comparison may help, and you can find licensed practitioners in the Gyfts practitioner directory or explore other approaches.