A Cochrane review ¹ includes 64 studies with a total of 7104 patients. Twenty-five trials were of 6 weeks' duration, 17 were of 6 and a half to 12 weeks' duration, and 17 were of less than 6 weeks' duration. Duration of treatment was unclear in 5 trials.

Acupuncture vs. no treatment/wait list/treatment as usual: Acupuncture (manual and electro) may reduce the severity of depression by end of treatment (SMD -0.66, 95% CI -1.06 to -0.25; 5 trials, n=488). It is unclear whether there are differences between groups in the risk of adverse events (RR 0.89, 95% CI 0.35 to 2.24; one trial, n=302).

Acupuncture vs. control acupuncture (invasive, non-invasive sham controls): Acupuncture may be associated with a small reduction in the severity of depression of 1.69 points on the Hamilton Depression Rating Scale (HAMD) by end of treatment (95% CI -3.33 to -0.05, 14 trials, n=841). It is unclear whether there are differences between groups in the risk of adverse events (RR 1.63, 95% CI 0.93 to 2.86; 5 trials, n=300).

Acupuncture vs. medication Acupuncture may confer small benefit in reducing the severity of depression by end of treatment (SMD -0.23, 95% CI -0.40 to -0.05; 31 trials, n=3127). Studies show substantial variation resulting from use of different medications and acupuncture stimulations. There may be lower ratings of adverse events following acupuncture vs. medication alone, as measured by the Montgomery-Asberg Depression Rating Scale (MADRS) (MD -4.32, 95% CI -7.41 to -1.23; 3 trials, n=481).

Acupuncture plus medication vs. medication alone: Acupuncture is highly beneficial in reducing the severity of depression by end of treatment (SMD -1.15, 95% CI -1.63 to -0.66; 11 trials, n=775). Studies show substantial variation resulting from use of different modes of acupuncture stimulation. It is unclear whether differences in adverse events are associated with different modes of acupuncture (SMD -1.32, 95% CI -2.86 to 0.23; 3 trials, n=200).

Acupuncture vs. psychological therapy: There were no differences between acupuncture and psychological therapy in the severity of depression by end of treatment (SMD -0.5, 95% CI -1.33 to 0.33; 2 trials, n=497). There are no differences between groups in rates of adverse events (RR 0.62, 95% CI 0.29 to 1.33; one trial, n=452).

Comment: The quality of evidence is downgraded by study quality (inadequate allocation concealment and lack of blinding), inconsistency (heterogeneity in treatments and controls) and indirectness (most of the studies were conducted in China, short follow-up time).