MIND-BODY MEDICINE:
COMPLEMENTARY THERAPIES FOR DEPRESSION — AN OVERVIEW

Edzard Ernst, MD, PhD, FRCP(Edin); Julia I. Rand, MBBS, MSc; Clare Stevinson, BSc, MSc

Depression is one of the most common reasons for using complementary and alternative therapies. The aim of this article is to provide an overview of the evidence available on the treatment of depression with complementary therapies. Systematic literature searches were performed using several databases, reference list searching, and inquiry to colleagues. Data extraction followed a predefined protocol. The amount of rigorous scientific data to support the efficacy of complementary therapies in the treatment of depression is extremely limited. The areas with the most evidence for beneficial effects are exercise, herbal therapy (Hypericum perforatum), and, to a lesser extent, acupuncture and relaxation therapies. There is a need for further research involving randomized controlled trials into the efficacy of complementary and alternative therapies in the treatment of depression.

(Arch Gen Psychiatry. 1998;55:1026-1032)

Depression is a frequently occurring psychiatric disorder with a prevalence of approximately 5% in the general population.1,2 It is estimated that at least one third of all individuals are likely to experience an episode of depression during their lifetime.3 Depression results in high personal, social, and economic costs through suffering, disability, deliberate self-harm, and health care provision. Despite the availability of drug and psychotherapeutic treatments, much depression remains undiagnosed or inadequately treated.4 This state of affairs has stimulated the development of educational campaigns and treatment consensus statements.5,6

Complementary and alternative medicine (CAM) is often negatively defined, for example, as "a system of health care which lies for the most part outside the mainstream of conventional medicine."7 A more inclusive definition8 has been adopted by the Cochrane Collaboration: "complementary medicine is diagnosis, treatment, and/or prevention which complements mainstream medicine by contributing to a common whole, by satisfying a demand not met by orthodoxy, or by diversifying the conceptual frameworks of medicine."

Complementary and alternative therapies (CATs) are popular. In 1991, 34% of the US adult population used at least 1 such therapy for 1 year.9 This figure has now risen to 40%.10 Twenty percent of those suffering from depression had used an unconventional therapy within the past year.9 Depression is among the 10 most frequent indications for using CATs, and relaxation, exercise, and herbal remedies are the 3 most prevalent CATs tried for this condition.10 Forty-two percent of 115 Danish psychiatric inpatients had used CATs at least once, with herbal medicine being the most frequent type.11 Herbal remedies, homeopathy, acupuncture, massage, relaxation, and unconventional psychotherapeutic approaches have been reported12 as the most prevalent CATs among psychiatric patients.

The response to the question of why people turn toward CAM is as fascinating as it is complex. No simple, uniform answer can be identified as the list of motivations will vary depending on which (patient) group one asks. Generally speaking, however, people opt for CAM because they want to leave no option untried and look for treatments devoid of adverse effects.13 Another important reason is that CATs are viewed as less authoritarian and more empowering, offering more patient conetrol.14 Astin10 found that CAM users are, on average, better educated and report poorer health than nonusers. Interestingly, they do not usually turn to CAM as a result of being dissatisfied with orthodox medicine, and only 5% use CAM as a true alternative to conventional medicine.

Lay books on CAM15-18 promote a wide range of CATs for depression (Table 1). In view of such promotion and CAM's popularity, the need for more information arises.21 The aim of this article, therefore, is to review the published evidence regarding the effectiveness of CAT in the treatment of depression. As we will see, the trial data are almost invariably burdened with numerous limitations. Small sample size, selection bias, uncertainty about the diagnosis, lack of blinding, lack of adequate outcome measures, failure to control for nonspecific therapeutic effects, failure to control for confounders, inadequate duration, and personal belief of the investigator in the treatment are the most frequent drawbacks.

Computerized literature searching without language restrictions was carried out to identify all randomized controlled trials (RCTs) relating to CATs used for depression. The following databases were searched: MEDLINE (literature from 1966-1996), EMBASE (literature from 1986-1996), CISCOM (Centralized Information Service for Complementary Medicine; search performed in January 1997), and the Cochrane Library (accessed March 1997 [Issue 1]). A wide range of search terms was used, reflecting the diversity of CATs: acupuncture, affective disorders, Alexander technique, alternative medicine, aromatherapy, art therapy, Bach (flower remedies), balneology, chiropractic, color therapy, complementary medicine, depression, depressive disorders, energy, essential oils, exercise, healing, herbal medicine, hydrotherapy, hypnosis, kinesiology, laughter, manipulation, massage, music, naturopathy, osteopathy, oxygen, polarity, qigong, reflexology, relaxation, therapeutic touch, and tragerwork. The reference lists of all articles thus found were also searched. Furthermore, inquiries were made to colleagues for any further publications and our files were searched.

Ideally, only RCTs would be selected for this review. However, as the search revealed that few RCTs have been conducted, less-rigorous studies are referred to in cases in which no RCTs are available. Articles discussing the following mainstream treatments for depression were excluded: cognitive therapy, light therapy for seasonal affective disorder, and partial sleep deprivation. Articles with no factual data were also excluded. All studies admitted to this review were read in full by two of us (E.E. and J.I.R.). Data were extracted according to a predefined checklist. Discrepancies were settled through discussion.

Acupuncture

Acupuncture is an ancient Chinese treatment. Based on the belief that 2 types of "energies" flow in "meridians" throughout the body and that an imbalance of these energies constitutes illness, acupuncturists insert needles into points located on meridians with the aim of correcting the imbalance and restoring health. Western acupuncturists are critical of these Taoist theories and attribute acupuncture's alleged benefits to neurophysiological effects.22 Hence, the putative mechanism for acupuncture in depression is provided through studies23 showing that the level of endorphins can be increased through needling. Acupuncture is normally carried out in specialized clinics either by physicians or (more often) by nonmedically qualified therapists (NMQTs). One session would typically last for 20 minutes, and a series of 6 to 12 treatments may be required. Case series24,25 indicate that acupuncture is promising for treating depression. Several uncontrolled26,27 and controlled28 clinical trials provide data in support.

Electroacupuncture appears to have greater efficacy than traditional acupuncture, and the preliminary results29 of a trial comparing standard electroacupuncture and computer-controlled electroacupuncture have been published. These indicate that the computer-controlled electroacupuncture treatment produced greater clinical improvement than electroacupuncture (P<.05) as measured by the grading system commonly used in China for the assessment of therapeutic effects.

Two RCTs30,31 compare the effects of electroacupuncture and amitriptyline hydrochloride in depressed patients. Patients suffering from depression (defined according to National Survey and Coordination Group of Psychiatric Epidemiology standards) were grouped at random to receive 5 weeks of therapy with either electroacupuncture (n=27) or the tricyclic antidepressant amitriptyline hydrochloride (n=20; average daily dose, 142 mg).30 A comparison of Hamilton Depression Scale scores before and after treatment showed a significant reduction (from 29 to 13 and 29 to 14, respectively) in the scores for both groups (P<.01). At the end of the treatment period, there was no statistically significant difference between the 2 groups.

An RCT31 involving 241 depressed patients compared treatment with electroacupuncture or amitriptyline hydrochloride for 6 weeks. Hamilton Depression Scale scores showed a significant reduction after treatment in both groups (from 35 to 8 and 35 to 10, respectively). There was no significant intergroup difference after 6 weeks. Follow-up of 148 patients for 2 to 4 years revealed no significant difference in the depression recurrence rate between the 2 groups.

Herbal Medicine

Medical herbalism (also termed phytotherapy in Europe) is the treatment of illness with plants, parts of plants, or plant extracts. It has a long history in all medical cultures, and many of our modern drugs have been derived from botanical sources. Each plant contains a whole array of compounds, and it is sometimes difficult to define which and how many of these contribute to which pharmacological effect. The mechanism of action can thus be complex, but may be understood or researched by conventional pharmacological methods. While the general public usually view plant-based medicines as devoid of adverse effects, this notion can be dangerously misleading.32 In continental Europe, phytotherapy is an integral part of physicians' prescribing. In the United States and the United Kingdom, herbal medicine is mostly in the hands of NMQTs. Scattered references33,34 occur in the ethnobotanical literature to plants used by indigenous peoples to treat depression. In China, herbal remedies are often used in combination with conventional western drug therapy.35 However, only few trials, usually of poor methodological quality, investigate Chinese herbal therapies for depression. A similar situation exists in Japan where traditional herbal mixtures are used for depression, but their effects have not yet been scientifically tested.36

Lay books on CAM15-18 claim a variety of plants to be helpful in depression, eg, wild oats, lemon balm, ginseng, wood betony, basil, and St John's Wort. Yet, only for St John's Wort (Hypericum perforatum) does a substantial body of evidence exist. It has recently been reviewed37,38 in English. The meta-analysis by Linde et al38 identified 23 RCTs involving a total of 1757 outpatients suffering from mild to moderate depression. Fifteen of these trials were placebo controlled, and 8 compared H perforatum with orthodox antidepressants. The overall responder rate ratio showed that H perforatum was significantly superior to placebo (2.67; 95% confidence interval, 1.78–4.01). H perforatum was found to have an efficacy similar to that of standard antidepressants. Compared with the antidepressant groups, the H perforatum groups had lower dropout rates (7.7% vs 4%) and numbers of patients reporting adverse effects (35.9% vs 19.8%). A recent comparative analysis (C. S. and E. E., unpublished data, June 1998) of adverse effects concluded that "Hypericum seems to be at least as safe and possibly safer than conventional antidepressant drugs."

Exercise

Many categories of physical exercise exist, eg, leisure-time and work-related physical activity or single bout and regular exercise. Their physiological responses may differ considerably. For the purpose of the following discussion, it is helpful to distinguish between regular endurance (mostly aerobic) exercise and power (mostly anaerobic) exercise. For the treatment of depression, exercise can be carried out either under supervision (eg, by a physiotherapist) or independently at home. In practice, a combined approach is usually the best.

A large body of evidence39 (>1000 trials) exists relating to exercise and depression and numerous reviews40-53 on the topic have recently been published. A meta-analysis of 80 studies50 (regardless of their methodological quality) produced an overall mean exercise effect size of –0.53 (range, –3.88 to 2.05). This suggests that the depression scores decreased by approximately one half of an SD more in the exercise groups than in the comparison groups. The antidepressant effect occurred with all types of regular exercise, independent of sex or age, and it increased with the duration of therapy. Overall, exercise was as effective as psychotherapy.

The available evidence suggests that any type of exercise alleviates depression. Martinsen and Stephens49 identified 8 experimental exercise-intervention trials in clinically depressed patients, and exercise was associated with reductions in depression scores in all of the studies. Two further RCTs54,55 were identified via our search strategy. In the first study54 moderately depressed elderly subjects were randomly allocated to walking exercises, social-contact control condition, or a waiting-list control group. After 6 weeks, the first 2 groups showed a significant decrease in Beck Depression Inventory scores compared with baseline. The second RCT55 involved 124 depressed subjects allocated to aerobic exercise, low-intensity exercise, or to a no exercise-intervention group. All subjects continued their usual psychiatric treatment. No significant difference was found in the Beck Depression Inventory scores between the groups after 12 weeks. However, the control group had been significantly more depressed at baseline.

Aromatherapy

Aromatherapists (normally NMQTs) use a combination of gentle massage techniques and essential oils from plants. These oils are thought to have specific pharmacological effects after transdermal resorption. One treatment would last about 30 minutes, and a series of 6 to 12 treatments would usually be recommended.

Although aromatherapy is advocated for improving mood in depression,56 and is perceived as helpful by some patients,57 there is very little objective evidence. In a small pilot study,58 12 depressed men were exposed to citrus fragrance in the air and compared with 8 patients who were not exposed to the fragrance. Both groups were taking antidepressants. It was reported that the dose of antidepressants in the experimental group could be markedly reduced. The study was not randomized and involved only a small number of patients with varying dose and type of antidepressants. At present, it is not possible to draw any firm conclusions about the value of aromatherapy for depression.

Dance and Movement Therapy

A dance therapist (usually an NMQT) aims to involve patients through encouragement to express themselves in movement and therefore enhance well-being. Treatments can be organized as group sessions, adding an additional element of social interaction. Typically, a session lasts 30 to 40 minutes, and regular (eg, weekly) repetitions are normally recommended.

Little scientific evidence is available for the role of dance and movement therapy.59 Only 2 studies60,61 were found, neither involving large numbers or of rigorous design. Twenty hospitalized psychiatric patients and 20 normal control subjects were divided into 4 groups.60 Half of the psychiatric patients and half of the controls received 1 dance and movement therapy session, and the other subjects received no intervention. After therapy, only the psychiatric patients showed a significant reduction in depression as measured by the Multiple Affect Adjective Checklist self-rating scale (P<.001). In the second study,61 12 inpatients with major depression were randomly assigned to movement therapy sessions on 7 of 14 days. Five of the patients showed a reduction in depression scores on movement therapy days compared with days with no therapy (P<.05). Both studies suffer from methodological limitations. Thus, insufficient evidence exists to assess the effect of dance and movement therapy in depression.

Homeopathy

Homeopathy is based on the "like cures like" principle that suggests that a remedy (often, but not always, plant based), which causes certain symptoms in a healthy individual, can be used as a treatment for patients presenting with such symptoms. Furthermore, homeopaths believe that, by "potentizing" (stepwise dilutions combined with vigorous shaking) a remedy, it will get not less, but more, potent. They assume that even dilutions devoid of molecules of the original remedy will have powerful clinical effects.62 Homeopathy is practiced by both physicians and NMQTs. A first consultation will usually last in excess of 1 hour.

There is a dearth of investigations into homeopathy for depression. The literature consists mainly of unsubstantiated treatment suggestions or case reports.63,64 The thorough review by Kleijnen et al65 and a recent meta-analysis by Linde et al66 of clinical trials of homeopathy detected only 1 study related to depression. It67 compared homeopathic treatment with diazepam in mixed anxiety and depressive states. This open trial was of low methodological quality, but produced a result in favor of homeopathy. A working group of the European Union located 377 reports of trials f homeopathy, which included no further studies in depression.68 The value of homeopathy as a treatment of depression is, therefore, presently unknown.

Hypnotherapy

Hypnotherapy is a state of focused attention or altered consciousness. All current theories of hypnosis are provisional and incomplete.69 Hypnotherapy cannot cure disease, but can be a useful adjunct to conventional treatments. Therapy sessions vary in length and rate of repetitions. Hypnotherapy is practiced both by physicians and NMQTs.

The literature on the subject consists only of anecdotal accounts and case reports.69,70 Our literature searches discovered no controlled clinical trials. It has been suggested71 that hypnotherapy may facilitate the process of cognitive therapy by aiding the restructuring of negative thought patterns. Again, this has not been substantiated. The value of hypnotherapy for depression is, therefore, not known at present.

Massage Therapy

There are several different forms and traditions of massage therapy.72 In the context of this article, massage uses typically a gentle manual stroking technique over the body (usually the back). This has a number of complex physiological and psychological effects, not least of which is relaxation of both the musculature and the mind.72 A treatment, usually carried out by an NMQT, would normally last for 20 to 30 minutes and a series of approximately 6 twice weekly sessions would constitute a typical prescription.

Most publications relating to massage and depression were found to consist of anecdotal accounts and case studies.73,74 A recent review75 of massage therapy uncovered only a few controlled trials. An RCT76 allocated 122 intensive care unit patients to receive either massage, massage with 1% lavender (Lavendula vera) oil, or rest periods. Those who received the massage with lavender oil reported a greater improvement in mood as measured by a self-rating 4-point scale. The study did not involve patients with depression, was short-term, and used a crude outcome measure. It is thus not possible to draw firm conclusions from its results.

In a well-conducted RCT,77,78 72 hospitalized children and adolescents, half with adjustment disorder and half with depression, either received 30-minute back massages (n=52) daily for 5 days or watched a relaxing video (n=20) for the same period. Profile of mood states depression scores were significantly lower immediately after massage compared with pretreatment values (P=.005). In addition, the premassage profile of mood states scores significantly declined during the 5-day treatment period (P=.01), and the massage group was less depressed than the control group at the end of the study. Because of the small sample size and the short treatment period, the data are insufficient to judge the value of massage for depression.

Music Therapy

Music therapy is the active or passive use of music to promote health and well-being. During treatment, patients perform music or listen to music carefully chosen and supervised by a trained music therapist (usually an NMQT). The type of music will depend on the personality and condition of the patient.

A limited amount of work relates to the effects of music therapy on depression.79 The results of an observational study80 using psychodynamic music therapy methods with depressed inpatients suggest that there may be a beneficial effect. One RCT81 involved 30 elderly patients (aged 61-86 years) with depression. They were randomly allocated to either a home-based music therapy program, a self-administered music therapy program, or a nonintervention waiting list (control group). After 8 weeks, the Geriatric Depression Scale scores of the 2 music groups were significantly better than those of the control group (P<.05). There is a need for further trials with larger numbers to determine whether this result can be replicated.

Relaxation Therapy

Relaxation therapy is an umbrella term for several techniques primarily aimed at decreasing physical and mental tension. Such treatments may include elements of meditation, yoga, and other mind-body therapies. They would normally be carried out by NMQTs.

Three RCTs82-84 investigating the effects of relaxation therapy were found. In the first study, 30 psychiatric outpatients with depression, all taking medication, were randomized to 3 groups.82 Two of the groups were given different forms of relaxation therapy during 3 days, while the third group acted as a control. Compared with controls, both relaxation-therapy groups showed a significant improvement in symptom scores (P<.05). However, a symptom score list was used that had not been validated, the sample size was small, and the treatment period short.

In an RCT83 involving 37 moderately depressed patients assigned to cognitive behavior therapy, relaxation therapy, or tricyclic antidepressants, the first 2 interventions resulted in significantly better mean Beck Depression Inventory scores than the pharmacological treatment (P<.01). The results should be viewed with caution because of the small sample size, lack of control for the nonspecific effects of attention from professionals, and reported noncompliance with the medication regime.

An RCT84 in 30 moderately depressed adolescents showed that relaxation training or cognitive behavior therapy resulted in a greater improvement than no intervention. Again, the sample size was small and there was no control for nonspecific effects.

On balance, therefore, relaxation treatments are promising, but further research and replications are required.

Conclusions

Because of the nature of the evidence relating to CAM and depression, a qualitative overview seemed preferable to a systematic review. Collectively, the above data suggest that exercise and H perforatum are effective symptomatic treatments for mild to moderate depression. The evidence for acupuncture, massage, and relaxation is promising, but not compelling.

Acupuncture and electroacupuncture can stimulate the synthesis and release of the monoamines serotonin and noradrenaline-norepinephrine in animals.85 This is the postulated mechanism for the perceived beneficial effect of acupuncture in depression. The evidence available on the efficacy of electroacupuncture in the treatment of depression has mainly come from 1 research group at the Institute of Mental Health, Beijing, China. The limited number of RCTs suggest a beneficial effect of a similar magnitude to that produced by amitriptyline hydrochloride. Electroacupuncture is reported to produce fewer and less-severe adverse effects than standard antidepressants. However, there is a need for the results to be replicated in rigorously designed RCTs using clear diagnostic criteria for patient entry, specified randomization procedures, and control for nonspecific responses resulting from the time and attention received during the acupuncture therapy.

Despite the potential of plant extracts as psychoactive substances, H perforatum is the only herb that has been investigated rigorously. The results show promising effects in patients with mild to moderate depression. However, they need to be followed up by further studies with more clearly defined diagnostic groups, groups of patients with major depression, standardized preparations, trials longer than 8 weeks, and comparison with antidepressant doses within the normal therapeutic range.86,87 H perforatum is associated with a markedly better adverse effect profile than standard antidepressants.88 This could lead to better compliance, quality of life, and efficacy.

The results of exercise-intervention studies indicate that there is an overall association between exercise and reduction in the symptoms of mild to moderate depression. However, many studies suffer from significant methodological flaws that make it difficult to draw firm conclusions.46 Many of the investigations are not of RCT design, involve only small numbers of subjects, are not controlled for the nonspecific effects of exercise, such as attention from trainers and social interaction where a group is involved, do not give full details of the exercise intervention, and use a variety of mainly self-reporting depression scales without objective blinded assessment. As with other CATs, it is unclear how long the antidepressive effects (if any) would persist.

A number of mechanisms by which exercise may improve mood have been proposed.41,46 These include physiological effects, such as changes in endorphin and monoamine levels; psychological effects, such as subject expectation, diversion from stressful stimuli, the effects of receiving attention, improved self-image, and feelings of control; and sociological factors, such as the benefits of social interaction and support. Although some longitudinal epidemiological evidence89,90 indicates that there may be a strong link between exercise and a reduction in depression levels, it is necessary to investigate this possibility further via high quality RCTs.

Few clinical studies are available regarding the effectiveness of other CATs in the treatment of depression. The data that do exist are generally of poor methodological quality. There are some indications that aromatherapy, massage, music therapy, and relaxation techniques may be of value. These areas thus warrant further investigation. No data exist regarding the efficacy of other therapies such as Alexander technique, Bach flower remedies, color therapy, kinesiology, naturopathy, polarity, tragerwork, qigong, and reflexology. In CAM, there is heated debate about which research methods might be appropriate. Some claim that this area of medicine is so different that it defies standard research methods. This, however, has repeatedly been demonstrated to be wrong (as shown by White et al91 and Vickers et al92). Clearly, the optimal method has to be chosen according to the research question and not to some vague ideological underpinning. If the question relates to testing the efficacy of a given treatment for depression, the RCT is unquestionably the design option that best excludes bias (eg, as summarized by Ernst,93 Sibbald and Roland,94 and Ernst95). In conclusion, apart from H perforatum and exercise, little rigorous scientific evidence exists regarding the effectiveness of CATs in depression. In view of the public's demand for CAT, investigation of these therapeutic options by well-designed RCTs is important.