Depression and Acupuncture: A Controlled Clinical Trial

Conclusions: In this study, based on a small outpatient sample of women with major depression, it appeared that acupuncture provided significant symptom relief at rates comparable to standard treatments such as psychotherapy or pharmacotherapy. The effect sizes observed in this small sample were at least as large or larger than those seen in trials of antidepressant medication or psychotherapy, and they suggest that a larger clinical trial is warranted.


by John J.B. Allen, Ph.D.

March 2000, Vol. XVII, Issue 3


Although acupuncture has been practiced for over 3,000 years (Ulett et al., 1998), the Western psychiatric scientific community has produced little empirical research on its efficacy. With the exception of investigations of acupuncture as a treatment for substance abuse and dependence (Brewington et al., 1994; McLellan et al., 1993), there are very few well-controlled empirical studies of acupuncture's efficacy for psychiatric disorders.

One difficulty with conducting well-controlled trials of acupuncture stems from an apparent conflict between controlled scientific research and the principles of Chinese medicine on which acupuncture is based. While Western scientific methods require rigorous controls, standardization of treatments and replicability, Chinese medicine requires flexibility and individualization in assessment and treatment planning. While some investigators have suggested that Chinese medicine, therefore, would require special research designs, our approach has been to conduct the gold standard, double-blind, randomized control trial of acupuncture to assess its efficacy for treating depression.

Previous Research

Aside from our study summarized in this article, the few studies of acupuncture as a treatment for depression or depression-like syndromes have been published in China (Han, 1986; Chengying, 1992), Eastern Europe and the former Soviet Union (Cherkezova and Toteva, 1991; Dudaeva et al., 1990; Frydrychowski et al., 1984; Poliakov, 1987). Since the diagnostic criteria used in these studies differed from those of the DSM-IV, and since most of these studies (other than their abstracts) have not been translated into English, it is difficult to fully evaluate them. Collectively, however, these studies suggest that acupuncture can be effective in the treatment of depression and depressive symptoms, and it may be as effective as tricyclic antidepressant medications in some cases. These findings encouraged us to undertake a well-controlled pilot study to examine the efficacy of acupuncture as a treatment for depression.

Methodological Concerns

Individualization of Treatments. In order to standardize a treatment delivery while preserving the flexibility and individual tailoring required by Chinese medicine, we developed a replicable, manual-based treatment approach (Schnyer et al., in press). Individuals who share the Western diagnosis of major depression are characterized by a variety of "patterns of disharmony" based on traditional Chinese medicine, each of which would dictate a particular treatment. Treatments, therefore, must be tailored to each individual's symptom picture; a standard treatment must not be uniformly administered to all individuals who fall within the Western diagnosis.

Our treatment manual (Schnyer et al., in press) describes the framework and procedures by which acupuncturists may reliably arrive at individualized diagnoses and treatment plans. Acupuncturists can agree on diagnoses and treatments using this manual; when using it, we obtained an intraclass correlation coefficient of 0.75, summarizing the composite agreement of four acupuncturists.

Blinding. A distinct advantage of using this individualized approach, as discussed later in this article, is that it facilitates the blinding of the acupuncturist providing treatment. Although a double-blind study is challenging to carry out, it is often assumed in research on mental disorders that the study is double-blind in two senses. The first is that neither the patient nor the clinician rating the outcome is aware of what treatment was received. The second is that neither the patient nor the provider knows what treatment was received. So, in a sense, the gold standard for treatment should entail a triple-blind study, where neither client nor provider nor outcome assessor is aware of the treatment received by the client.

When it comes to conducting efficacy trials of acupuncture in Western cultures, blinding the patient is relatively easy, as long as the patient perceives some needling. Few individuals in Western society know where acupuncture points are located, much less the points that might aid in their particular presentation of depression. It is much more difficult to blind the treatment provider. Using inactive (sham) acupuncture does not adequately blind the treatment provider, as the provider is fully aware which treatments are valid and which are invalid. Such awareness will lead the provider to have different expectations of outcome for sham versus active treatments. These expectations can, and should, be assessed in any efficacy study.

The thorny issue of blinding the acupuncture treatment provider is difficult, but not impossible, to resolve. Because any two individuals with depression will rarely present with identical symptoms, any two depressed individuals are likely to receive rather different constellations of acupuncture points in their treatment. Thus, if any acupuncture treatment provider simply received a set of points to administer (and no information on specific symptoms or patient history), it would not be immediately obvious to that provider whether those points were designed to address an energetic imbalance underlying a particular client's depression. Following this logic a bit further, it is therefore possible to separate two traditionally integrated functions of the acupuncturist treatment provider: assessment and treatment. If assessors only served to conceptualize the treatment principles and devise the treatment strategy and associated points, and a different group of treating acupuncturists administered that strategy using those points, then there is a reasonable chance that these treating acupuncturists would be blind as to whether a treatment would be the most effective for a given client. Thus, while the providers would not be blind as to what points were used, they may be blinded as to the particular intent of the treatment.

This strategy presumes two important restrictions: 1) treating acupuncturists are prohibited from using their standard assessment procedures (including interview, palpation, taking of pulses and examination of the tongue); and 2) treating acupuncturists are not fully aware of the particular theoretical framework (i.e., combination of perspectives within Chinese medicine) being used to conceptualize the patient. While this strategy would not guarantee that the treating acupuncturist would remain blind, the effectiveness of this strategy in producing a double-blind can be monitored by simply administering questionnaires that tap the expectancies and beliefs of the provider and recipient. This is the strategy we used in our pilot study (Allen et al., 1998).

Controlling for Nonspecific Factors. Factors that are not specific to acupuncture or the points selected for treatment also have a therapeutic impact. Such factors include the patient-acupuncturist relationship and the engagement in activity believed to improve depression.

Therefore, our study design provided for the development of two types of acupuncture treatments for each patient: 1) specific treatment, or a treatment individually tailored to treat the patient's specific symptoms of depression; and 2) nonspecific treatment, or a treatment designed to treat a pattern of disharmony not related to the individual's depression but characteristic of the individual (e.g., targeting back pain). Specific and nonspecific treatments were similar from the patients' perspective, each involved points in the same general body regions. Moreover, patients were unaware of which treatment they were receiving. If specific treatments demonstrated greater efficacy than nonspecific treatments, then the effect of acupuncture per se is presumed to be responsible.

A Pilot Study

Our pilot study (Allen et al., 1998) was designed to examine the effectiveness of acupuncture as a treatment for major depression in women. A randomized clinical trial with blind outcome ratings assessed women with major depression who were randomly assigned to one of three treatment groups for eight weeks. Specific treatment involved acupuncture treatments for symptoms of depression; nonspecific treatment involved acupuncture treatment for symptoms that were not clearly part of the depressive episode. A wait-list condition, where patients waited without treatment for eight weeks, was also used. Nonspecific and wait-list conditions were followed by crossover to specific treatment.

A community volunteer sample was used with 33 women who met DSM-IV criteria for current major depressive episode of less than two years duration and who did not have other Axis I psychopathology. Thirty-eight women between the ages of 18 and 45 were recruited through newspaper advertisements that mentioned treatment for depression, but not acupuncture. Five women (13%) terminated prior to completion of the study. This resulted in a final sample of 33 women who received treatment specifically for depression.

The specific and nonspecific treatment plans were developed by an assessing acupuncturist and were administered by four board-certified acupuncturists, not including the assessing acupuncturist. Because the nonspecific treatments involved valid acupuncture points, treating acupuncturists perceived that they were providing a valid treatment-a belief that they would not have held if sham points had been used as a control. The treating acupuncturists were blind to experimental hypotheses and were not told which treatment plan they received. Nonetheless, that did not guarantee that this was a double-blind study. It remained possible that the acupuncturists developed some awareness of the differences between the treatments. On the other hand, the acupuncturists rated their beliefs about the efficacy of the treatment following each patient's first session; these ratings did not differ between specific and nonspecific treatments (F[1,22]<1, ns).

Results and Conclusions

Following treatments specifically designed to address symptoms of depression, 64% of women experienced full remission according to DSM-IV criteria. Comparing the immediate effect of the three eight-week treatment conditions, patients receiving specific acupuncture treatments demonstrated significantly (p<0.05) greater reduction in Hamilton Rating Scale for Depression scores (-11.7±7.3) than those receiving the nonspecific acupuncture treatments (-2.9±7.9), and showed marginally (p<0.12) more improvement than the wait-list controls (-6.1±10.9). Moreover, when examining the effect sizes, the specific treatment had a very large effect size compared to nonspecific treatment (d=1.16) and a moderate to large effect size compared to wait-list (d=0.61).

Thus, based on a small outpatient sample of women with major depression, it appeared that acupuncture provided significant symptom relief at rates comparable to standard treatments such as psychotherapy or pharmacotherapy. The effect sizes observed in this small sample were at least as large or larger than those seen in trials of antidepressant medication or psychotherapy, and they suggest that a larger clinical trial is warranted.

Dr. Allen is associate professor of psychology, cognitive science and neuroscience at the University of Arizona. In addition to clinical trials for acupuncture in the treatment of depression, his research focuses on electrophysiological and psychophysiological measures of emotion and risk for emotional disorders.


Allen JJB, Schnyer RN, Hitt SK (1998), The efficacy of acupuncture in the treatment of major depression in women. Psychological Science 9(5):397-401.

Brewington VB, Smith MS, Lipton DL (1994), Acupuncture as a detoxification treatment: an analysis of controlled research. J Subst Abuse Treat 11(4):289-307.

Chengying Y (1992), Mind-regulating acupuncture treatment of neurosis, using points of Du channel. International Journal of Clinical Acupuncture 3(2):193-196.

Cherkezova M, Toteva S (1991), [Reflexotherapy in the treatment of alcoholics with depressive syndrome.] Zh Nevropatol Psikhiatr Im S S Korsakova 91(2):83-84.

Dudaeva KI, Mikhailova ES, Poliakov SE (1990), [Neurophysiologic changes during the treatment of endogenous depression by reflexotherapy]. Zh Nevropatol Psikhiatr Im S S Korsakova 90(4):99-103.

Frydrychowski A, Landowski J, Watrobski Z, Ostrowska B (1984), [Use of acupuncture in the treatment of depressive syndromes]. Psychiatr Pol 18(3):247-250.

Han JS (1986), Electroacupuncture: an alternative to antidepressants for treating affective diseases? Int J Neurosci 29(1-2):79-92. McLellan AT, Grossman DS, Blaine JD, Haverkos HW (1993), Acupuncture treatment for drug abuse: a technical review. J Subst Abuse Treat 10(6):569-576.

Poliakov SE (1987), [Acupuncture in the treatment of patients with endogenous depression.] Zh Nevropatol Psikhiatr Im S S Korsakova 87(4):604-608.

Schnyer RN, Allen JJB, Hitt SK, Manber R (in press), Acupuncture in the Treatment of Major Depression: A Manual for Research and Practice. University of Arizona Press.

Ulett GA, Han J, Han S (1998), Traditional and evidenced-based acupuncture: history, mechanisms, and present status. South Med J 91(12):1115-1125.

PDF Version

Share this