Acupuncture for Snoring and Sleep Apnea: Reading a Large Positive Review Carefully
A network meta-analysis of 43 randomised trials and 3,402 patients concluded acupuncture is effective and safe for sleep apnea. There are three good reasons to hold that conclusion loosely, and they are worth understanding.
Acupuncture for sleep apnea is not a fringe question with no data behind it. There is a substantial randomised literature, and a recent synthesis of it is clearly positive.
It is also a good case study in why "43 randomised trials" is not the end of an argument.
What the review found
A 2025 network meta-analysis searched eight databases — PubMed, Web of Science, Cochrane, Embase, plus the Chinese databases Wanfang, CNKI, CBM and VIP — and pooled 43 randomised trials covering 3,402 patients with sleep apnea syndrome.
It ranked different acupuncture approaches rather than treating them as one thing:
| Outcome | Best-ranked approach | SUCRA |
|---|---|---|
| Reducing AHI | Electroacupuncture | 86.0% |
| Shortening longest apnea time | Acupoint catgut embedding | 98.7% |
| Lowest oxygen saturation | Acupoint catgut embedding | 89.7% |
| Daytime sleepiness | Manual acupuncture + Chinese herbal decoction | 87.6% |
| Overall effectiveness | Manual acupuncture + Western medicine | 79.8% |
| Fewest adverse events | Herbal decoction alone | 93.4% |
The authors' conclusion: acupuncture therapies are effective and safe for sleep apnea syndrome.
Three reasons to hold that loosely
The response threshold was lenient. "Effectiveness" was defined as a greater than 25% decrease in AHI plus a greater than one-third reduction in symptom scores. A 25% AHI reduction is a real change and it is a long way from treatment. Someone going from an AHI of 40 to 30 meets that bar while remaining severely apneic. Compare it with the definitions used elsewhere — hypoglossal nerve stimulation trials use AHI under 20 and a 50% reduction, and still describe that as response rather than cure.
SUCRA values are rankings, not effect sizes. A SUCRA of 86% means a treatment ranked highly across the network's probability distribution. It does not tell you how many events per hour anybody's AHI fell by. A therapy can rank first in a field of weak options.
Trial provenance matters here more than usual. Four of the eight databases searched were Chinese-language, and acupuncture trials conducted in China have a well-documented tendency to report positive results — a pattern identified across the acupuncture literature generally, not a criticism of any individual study. When nearly all trials of an intervention come from settings where almost no trial of it has ever been negative, the pooled estimate inherits that.
None of this means acupuncture does nothing. It means the honest reading is weaker than the abstract's conclusion.
Evidence tier 2. A large randomised literature exists and points positive. The response definition is loose, the pooled outputs are rankings rather than magnitudes, and the provenance pattern is a recognised source of inflation. That combination supports "plausible, possibly helpful" rather than "effective."
How to think about trying it
It is not a replacement for diagnosis or for treating moderate-to-severe apnea. No reading of this literature supports using acupuncture instead of CPAP for severe disease. If your AHI is high, the thing with large randomised effect sizes is positive airway pressure.
The risk profile is genuinely favourable, which is the strongest practical point in acupuncture's favour. Adverse events were a measured outcome in this review and they were low. Compared with surgery, that is a meaningful asymmetry — a low-risk intervention with uncertain benefit sits differently from a high-risk one.
If you try it, measure something. Snoring recordings before and after, or ideally a repeat sleep study if you have a diagnosis. The whole difficulty with this literature is soft outcomes, and you can avoid reproducing that problem in your own case. How to record your snoring properly.
Compare it against the exercise-based options first, because they target the same airway musculature with a clearer mechanism and better-characterised effect sizes. Myofunctional therapy for sleep apnea.
And apply the same scepticism to the positive studies you apply to the negative ones. Reading a single flattering abstract and stopping is how most bad remedy decisions get made. How to read sleep research in the news, and what we skip and why.
If you have not been assessed: witnessed breathing pauses, gasping arousals, waking unrefreshed, or sleepiness more sleep does not fix all point at a sleep study rather than any remedy. Snoring versus apnea.