Depression trials of acupuncture measure severity on rating scales such as the clinician-scored Hamilton Depression Rating Scale (HAMD), and compare acupuncture with no treatment, with sham acupuncture, with antidepressants, or with talking therapy. Cochrane’s 2018 review by Caroline Smith and colleagues found the difference from acupuncture shrank as the control became more like acupuncture: larger against no treatment than against sham. Its estimates of benefit were all rated low or very low quality. Sham acupuncture is a control procedure built to resemble acupuncture; Sham acupuncture covers the kinds in use and why none is inert.
The 2018 Cochrane review
This second update searched to June 2016 and included 64 randomised trials with 7,104 adults diagnosed with depression under standard criteria, including the Chinese classification CCMD-3-R. It covered manual acupuncture, electroacupuncture and laser acupuncture. The authors rated each result with GRADE, the scale Cochrane uses to say how much confidence a result supports, from high to very low. A standardised mean difference (SMD) expresses the gap between groups in units of the spread of results, so that trials using different scales can be pooled; a negative value favours acupuncture.
| Comparison | Trials (people) | Difference | GRADE |
|---|---|---|---|
| Acupuncture vs no treatment, waiting list or usual care | 5 (488) | SMD −0.66 | Low |
| Acupuncture vs sham acupuncture | 14 (841) | 1.69 points lower on the HAMD | Low |
| Acupuncture vs antidepressant medication | 31 (3,127) | SMD −0.23 | Very low |
| Acupuncture plus medication vs medication alone | 11 (775) | SMD −1.15 | Very low |
| Acupuncture vs psychological therapy | 2 (497) | SMD −0.5, confidence interval crossing zero | Low |
Most trials were at high risk of performance bias, which arises when the people giving or receiving a treatment know which one it is, and at high or unclear risk of detection bias, which arises when those assessing the outcome know. The comparisons with medication varied with the class of drug and the type of acupuncture stimulation. The authors could not say whether acupuncture differed from any control in adverse events, because most trials did not report them adequately. Few trials followed participants after treatment or measured quality of life. The authors concluded that high-quality trials are urgently needed against sham, medication and psychological therapy. After a treatment covers what acupuncture’s adverse events are and how often they occur.
Later reviews
A 2019 meta-analysis by Armour, Smith and colleagues narrowed the question to major depressive disorder and searched English, Chinese and Korean databases to November 2018. It included 29 trials with 2,268 participants, 22 of them run in China. Using Hedges’ g, another standardised measure of the difference between groups, it found reductions in severity against usual care (0.41), against sham (0.55) and when acupuncture was added to antidepressants (0.84). Severity fell further as the number of sessions rose. The authors flagged that most trials were at high risk of bias from lack of blinding, that Chinese trials gave more sessions more often, so their results may not transfer to other settings, and that few trials followed people up or reported safety well.
A 2024 meta-analysis by Tan and colleagues compared acupuncture alone with antidepressant drugs, using trials published in Chinese and English between July 2013 and July 2023: 20 trials with 1,376 participants. HAMD scores were lower with acupuncture after four weeks of treatment, with high inconsistency between trials (I² 86%), and side-effect scores were lower in the acupuncture groups.
Two trials
The ACUDep trial, published in 2013 by Hugh MacPherson and colleagues at the University of York, randomised 755 people with ongoing depression from 27 general practices in the North of England. Each was assigned to acupuncture (302), counselling (302) or usual care alone (151), and attended a mean of ten acupuncture or nine counselling sessions. The outcome was the PHQ-9, a nine-question depression score. At three months the score was 2.5 points lower with acupuncture and 1.7 points lower with counselling than with usual care; over 12 months the differences were 1.6 and 1.5. Acupuncture and counselling did not differ significantly from each other. The authors state the trial was not designed to separate the effects specific to needling from the rest of the encounter. No serious treatment-related adverse events were reported.
A 2011 trial at the University of Pittsburgh Medical Center compared electroacupuncture with a sham control: 53 adults with mild or moderate major depressive disorder had 12 sessions over six to eight weeks of either two-point electroacupuncture or needling at scalp points off the channels with sham electrical stimulation. HAMD scores fell by 6.6 points with electroacupuncture and 7.6 with the control, and the authors could not show a specific effect of electroacupuncture. Reading an acupuncture trial explains why a pragmatic trial like ACUDep and a sham-controlled one like this ask different questions.
What guidelines say
NICE’s guideline on depression in adults, NG222, was published in June 2022. In its rationale for treating a new episode of more severe depression, the committee found some evidence of effectiveness and cost effectiveness for acupuncture combined with antidepressants, but noted that the evidence came from Chinese acupuncture, which it considered different from Western acupuncture and possibly not applicable to people in the UK. The committee therefore made no practice recommendation and instead made a research recommendation, asking how effective and cost effective acupuncture plus antidepressants is for more severe depression in the UK.
The Canadian Network for Mood and Anxiety Treatments (CANMAT) reached a different position in its 2023 update. Citing several meta-analyses, and the Cochrane finding that the difference against sham was smaller than clinically relevant thresholds, it lists acupuncture as a second-line treatment for mild major depressive disorder and a second-line add-on treatment for moderate major depressive disorder. It found insufficient evidence for more severe illness. Clinical guidelines compares other guideline bodies.
What the reviews could not settle
The comparison with sham, which isolates needling, is small and rated low quality, and the one sham-controlled trial above found no difference. The comparisons with drugs and with acupuncture added to drugs come mostly from trials in China, rated very low quality, and both NICE and the 2019 review question how far they transfer elsewhere. The evidence on harms is thin because most trials did not report them adequately, and few trials looked beyond the end of treatment. The reviews report averages across trials; whether any treatment suits one person with depression is a question for a clinician who knows them.
AcuiQ’s depression page, with the related major depression and post-stroke depression pages, lists the protocols individual studies prescribed, each with its citation. Anxiety and Insomnia cover two conditions that appear alongside depression in some trials.