Long-term back and neck pain are among the most common reasons people try acupuncture. The research here is bigger than for almost any other use, but it needs reading carefully. Acupuncture does better than no acupuncture, and a little better than sham acupuncture, and the size of that second difference is what deserves a frank conversation.
What an individual patient data meta-analysis is
Most reviews combine the published summary results of trials. An individual patient data meta-analysis goes a step further. The researchers collect the raw data for every single participant in each trial and analyse it all together, which makes it one of the most thorough ways of pooling evidence.
A group called the Acupuncture Trialists' Collaboration, with Andrew Vickers as lead author, did exactly this for chronic pain.
The 2012 analysis
The first analysis, published in Archives of Internal Medicine in 2012, included 29 randomised trials with 17,922 patients who had chronic back and neck pain, osteoarthritis, chronic headache or shoulder pain.
Acupuncture beat both sham acupuncture and no acupuncture for every one of these conditions. For back and neck pain, once a group of trials with unusually strong results was set aside, pain scores were 0.23 standard deviations lower than sham, and 0.55 standard deviations lower than no acupuncture.
Those numbers are hard to picture, so the authors gave a worked example. Imagine a group starting with pain of 60 out of 100. If "success" means halving your pain, around 30% of people might get there with no acupuncture, 42.5% with sham acupuncture, and 50% with real acupuncture.
The 2018 update
A 2018 update in the Journal of Pain widened this to 39 trials and 20,827 patients. The results were much the same, with effects of roughly 0.5 standard deviations compared with no acupuncture and 0.2 compared with sham. The benefits also largely lasted, with only about a 15% drop in the treatment effect after a year.
So how big is the effect?
It depends what you compare it with.
- Compared with no acupuncture, the difference is moderate and worth taking seriously.
- Compared with sham, the difference is small. Much of the improvement also happens with sham needling.
The authors concluded that the pain relief "cannot be explained solely in terms of placebo effects", but also that factors beyond the needling itself are important contributors. The paper doesn't pin these down, though they may include things like time, attention and the setting of treatment. In 2012 the researchers called acupuncture "a reasonable referral option" for chronic pain.
Where NICE stands
The NICE guideline on low back pain (NG59, 2016) goes the other way. "Do not offer acupuncture for managing low back pain with or without sciatica," it states. A later guideline on chronic primary pain (NG193, 2021) allows a single course of acupuncture to be considered for pain with no clear underlying cause. Experts clearly read this evidence in different ways, and it's only fair that you know that.
If you're thinking about it
Persistent back or neck pain should be checked by your GP or a physiotherapist, especially if it's new, getting worse, or comes with other symptoms. If you then decide to try acupuncture, I'll be straightforward about what the research shows and agree a short course with you before reviewing how it's going.
You can read more about the studies I draw on in my research section.
I've also written dedicated pages on acupuncture for back pain and acupuncture for neck pain, with what a session involves.