Doi:10.1016/j.jad.2006.06.02

Journal of Affective Disorders 97 (2007) 23 – 35 A.F. Thachil ⁎, R. Mohan 1, D. Bhugra 2 Kings College London, Section of Cultural Psychiatry, HSRD, PO: 25, Institute of Psychiatry, DeCrespigny Park, London SE5 8AF, UK Received 31 January 2006; received in revised form 22 June 2006; accepted 23 June 2006 Background: Depression is one of the leading indications for using Complementary and Alternative Medicine (CAM). This paperreviews the evidence of efficacy of different types of CAM in depression with the aim of identifying the highest level of evidence.
Methods: We conducted literature searches restricted to the English language for studies on CAM as monotherapy in depression. Allpapers were reviewed by two researchers and the evidence was ranked according to a widely referenced hierarchy of evidence.
Results: 19 papers formed the final review. We found Grade 1 evidence on the use of St. John's wort, Tryptophan/5-Hydro-xytryptophan, S-adenosyl methionine, Folate, Inositol, Acupuncture and Exercise in Depressive disorders, none of which wasconclusively positive. We found RCTs at the Grade 2 level on the use of Saffron (Herbal medicine), Complex Homoeopathy andRelaxation training in Depressive disorders, all of which showed inconclusive results. Other RCTs yielded unequivocally negativeresults. Studies below this level yielded inconclusive or negative results.
Limitations: Searches were restricted to the English language. Our list of CAM approaches may not have been comprehensive. Weexcluded studies on the use of CAM as an adjunctive treatment and this review aimed to identify only the highest level of evidence.
Conclusions: None of the CAM studies show evidence of efficacy in depression according to the hierarchy of evidence. The RCTmodel and the principles underlying many types of CAM are dissonant, making its application in the evaluation of those types of CAMdifficult. The hierarchy of evidence we used has limited utility in grading trials of CAM.
2006 Elsevier B.V. All rights reserved.
Keywords: CAM; Complementary therapy; Depression; Depressive disorders; Evidence; Hierarchy of evidence Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
⁎ Corresponding author. Tel.: +44 207 837 8888x5615; fax: +44 207 277 1462.
E-mail addresses: (A.F. Thachil), (R. Mohan), (D. Bhugra).
1 Tel.: +44 207 848 0048; fax: +44 207 277 1462.
2 Tel.: +44 207 848 0047; fax: +44 207 277 1462.
0165-0327/$ - see front matter 2006 Elsevier B.V. All rights reserved.
doi: A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 Grade 1 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264.1.
Hypericum perforatum (St. John’s Wort) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 Tryptophan/5-Hydroxytryptophan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 S-adenosyl methionine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Acupuncture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Folate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Inositol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Grade 2 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295.1.
Saffron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 Homoeopathy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Relaxation training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Docosahexaenoic acid, tyrosine and Lavandula angustifolia Mill . . . . . . . . . . . . . . . . . . . . . . . . 30 Grade 3 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 306.1.
Music therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Yoga (Sudarshan Kriya) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 Movement therapy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Aromatherapy massage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Grade 4 and Grade 5 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Discussion. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Grade 1 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Grade 2 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 Grade 3 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Grades 4 and 5 evidence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 disability ). Taking the disabilitycomponent of burden alone, depression is the leading The last decade has witnessed a significant growth of cause of disability in both sexes and across all ages, interest in Complementary and Alternative Medicine accounting for almost 12% of all disability ( (CAM) worldwide. In the UK, about 1 in 10 of the adult population consults a CAM practitioner every year, and depression is under-diagnosed and under-treated, and 90% of these contacts happen outside the National compliance with antidepressants is often low 40% of the respondents of a national survey in the United nificance given the prevalence of CAM use by those who States had used at least one such therapy in the previous are depressed. The 3 most widely used CAM approaches 1 year (). In 1996, it was estimated that in depression are Relaxation, Exercise and Herbal almost half of Australian adults used complementary medicines in the past year and a fifth consulted Some of the reasons for the use of CAM include the relatively lower incidence of adverse effects, perceived 42% of 115 Danish psychiatric in-patients had used effectiveness, the desire for egalitarian relationships with Complementary and Alternative Medicine (CAM) at medical practitioners, a holistic approach to the indivi- least once during the course of their illness ( dual's problems and dissatisfaction with conventional Depression is one of the 10 most frequent indications number of CAM consultations, providers of conventional Depression is a significant cause of morbidity and mor- healthcare have failed to address the issue. The United tality world-wide, imposing a range of costs on indi- Kingdom Department of Health (DOH) has published viduals, families and communities and accounting for a sizeable proportion of the global burden of disease and mary Care Groups (PCGs) which deals with 6 therapies; A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 Acupuncture, Aromatherapy, Chiropractic, Homeopathy, Open study, Case series and Case study as search terms.
Hypnotherapy and Osteopathy. This has in part been We also hand searched for cross-referenced articles and responsible for several of these therapies being made books. The searches were restricted to publications in the accessible to patients by PCGs and the UK Department of Health funding several CAM research projects, as per the We obtained the list of CAM approaches from di- recommendations of the House of Lords Select Commit- verse sources—textbooks, primers and guides in the based on popularity and currently available evidence of efficacy. This is still a positive step, as it can provide and discussions. We supplemented this search with impetus for further research into these approaches, but this interviews with CAM practitioners in London and other may mean that those therapies that are not listed may be cities in the UK, to identify various forms of CAM. The final list included 55 Complementary and Alternative Given their frequent use, CAM approaches warrant the same level of evaluation as conventional treatments.
Studies that investigated the use of CAM as an adjunct Service users, planners, general practitioners and mental to conventional treatment were excluded from the review, health professionals need to be informed about which since we focussed on CAM as monotherapy. We also treatments are effective, which are not, and which ones excluded studies that evaluated a combination of two or have been adequately evaluated. With this goal in mind, more CAM approaches as the primary therapeutic we aimed to review studies that looked at the efficacy of intervention. Studies on subjects younger than 18 years different types of CAM in Depressive disorders among were excluded from the review. Studies that addressed adult patients, with the specific aim of identifying the depressive states outside ICD (International Classification highest level of available evidence for each of these of Diseases) and DSM (Diagnostic and Statistical Manual CAM approaches. For this purpose, we used a hierarchy of Mental Disorders) based diagnoses of depressive of evidence which is one of the most widely used disorders were excluded to ensure homogeneity.
frameworks for ranking research-based clinical knowl-edge, and underpins many health policy and managementdecisions ). The reviewed research was graded according to the levels of this hierarchy.
List of complementary and alternative therapies We hypothesised that many of these therapies would not meet the criteria for the best evidence i.e. levels 1 or 2 We conducted computerized literature searches to identify all studies related to the use of CAM as mono- therapy in Depressive disorders. The following databases were searched-MEDLINE (literature from 1966–June 2005), EMBASE (literature from 1980–June 2005), AMED (Allied and Complementary Medicine, literature from 1985–June 2005), PSYCINFO (1974–June 2005), the Cochrane Complementary Medicine field and the Cochrane Library (up to June 2005). Age restrictions were not applied. The search terms were Complementary Medicine, Complementary Therapy, Alternative Medi- cine, Alternative Therapy, Depression and Depressive Disorder. We combined these individually with System- atic Review, Meta-analysis, RCT, Case-control study, Cohort study, Open study, Case series and Case study for each search. Subsequently we also searched for individual therapies by name, combined with Systematic Review, Meta-analysis, RCT, Case-control study, Cohort study, A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 Two researchers independently reviewed the citations The studies we found will be discussed under the corre- retrieved from the search. One of these was the retrieving researcher. We identified potentially relevant abstracts,obtained the original articles and reapplied the inclusion and exclusion criteria. They were then critically ap-praised independently by each researcher for the Our search found Grade 1 evidence on the use of St.
methodology and the level of evidence, and were then John's wort, a herbal medicine, a number of Nutritional ranked according to the hierarchy of evidence Therapy approaches, Acupuncture and Exercise in The appraisal included assessment of sampling methods, Depressive disorders. None of this evidence was randomisation, blinding, statistical analyses, outcome unequivocally positive. Where the results tended towards measures, duration of follow-up, monitoring of side the positive, the evidence was not sufficiently conclusive effects and criteria used in systematic reviews. Disagree- ments were resolved by discussion. For those therapieswhere evidence at a particular level was available, we did 4.1. Hypericum perforatum (St. John’s Wort) not search for evidence below that level e.g. wheresystematic reviews or randomised controlled trials were We found Grade 1 evidence for the use of Hypericum available, we did not search further for evidence at lower Perforatum (St. John's Wort) in depressive disorders levels, which was lower in the hierarchy of evidence.
). However, results of placebo-con-trolled trials showed marked heterogeneity. In trials restricted to patients with major depression, the com-bined response rate ratio (RR) for hypericum extracts We identified 59 studies that evaluated the use of compared with placebo from six larger trials was 1.15 CAM in depressive disorders. Of these, we excluded 12 (95% confidence interval, 1.02–1.29) and from six systematic reviews since they pre-dated the latest in that smaller trials was 2.06 (95% CI, 1.65 to 2.59). In trials particular CAM; one systematic review because it stud- not restricted to patients with major depression, the RR ied anxiety in depressed patients, and 12 papers since from six larger trials was 1.71 (95% CI, 1.40–2.09) and they were non-systematic reviews or commentaries.
from five smaller trials was 6.13 (95% CI, 3.63 to 10.38).
Besides these, 16 more papers were excluded as they had Compared with selective serotonin reuptake inhibitors either one or more of the following methodological (SSRIs) and tri- or tetracyclic antidepressants, respec- problems: they (i) evaluated the use of CAM as an tively, RRs were 0.98 (95% CI, 0.85–1.12; six trials) and adjunct to conventional therapy, (ii) evaluated the use of 1.03 (95% CI, 0.93–1.14; seven trials). Patients given St.
a combination of complementary therapies, (iii) did not John's wort dropped out of trials due to adverse effects use operationalized (ICD or DSM) diagnoses of less frequently than those given older antidepressants depression, (iv) studied samples with mixed diagnoses (Odds Ratio: 0.25; 95% CI, 0.14–0.45). Patients given (v) evaluated the use of CAM in subjects under 18 years, St. John's wort dropped out of trials due to adverse (vi) contained little information about the methodology; effects less frequently than those given SSRIs, but this and (vii) did not have definite outcome measures. 19 difference was not statistically significant (OR: 0.60, papers formed the final review and are shown in We found Grade 1 evidence indicating some benefit Grade Strong evidence from at least 1 systematic review of multiple from Nutritional Therapy in Depression, with Trypto- well designed randomised controlled trials.
Grade Strong evidence from at least 1 properly designed randomised controlled trial of appropriate size.
systematic review found only 2 trials, out of 108 ex- Grade Evidence from well-designed trials without randomisation, tracted from literature, to be of sufficient quality to meet single group pre-post, cohort, time series or matched case- the inclusion criteria. They involved a total of 64 patients and the available evidence suggests these substances Grade Evidence from well-designed, non-experimental studies from more than one centre or research group.
were better than placebo at alleviating depression (Peto Grade Opinions of respected authorities, based on clinical evidence, Odds Ratio 4.10; 95% confidence interval 1.28–13.15; descriptive studies or reports of expert committees.
RD 0.36; NNT 2.78). However, the evidence was of insufficient quality to be conclusive. The possible A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 Table 3Highest level of evidence for CAM approaches in depressive disorders Minimal beneficial effect, with a Response Rate Grade 1, ratio (RR) of 1.15; (95% CI, 1.02–1.29) over equivocal.
placebo in 6 larger trials and 2.06 in 6 smallertrials (95% CI, 1.65–2.59).
Significant effect over placebo (Peto odds ratio Grade 1, 5-hydroxy-tryptophan for depression.
4.10; 95% CI 1.28–13.15; RD 0.36; NNT 2.78), equivocal.
but sample size small (64), duration short (10weeks) and data insufficient to evaluate adverseeffects.
Greater response rate with SAMe compared to Grade 1, placebo (global effect size 27–38% depending equivocal.
on definition of response), antidepressant effectcomparable to standard tricyclics. Most studieswere methodologically flawed and brief induration (b6 weeks).
Results from 5 trials Included in the meta- Grade 1, analysis showed no difference in the reduction equivocal.
in the severity of depression compared tomedication (WMD 0.53, 95% CI −1.42 to2.47). Evidence insufficient to determine theefficacy of acupuncture vs. medication due tothe poor methodological quality of the trials.
Standardised mean difference in effect size −1.1 Grade 1, (95% CI, −1.5 to −0.6) over no treatment. No equivocal.
statistically significant difference in effect sizeover cognitive therapy. All studies had importantmethodological weaknesses.
Nutritional therapy—Folate for depression.
No significant benefit when used instead of Grade 1, antidepressant. Significant effect with folate equivocal.
augmentation (avg. reduction in HDRS scoresof 2.65 points, 95% CI 0.38–4.93; NNT for 50%reduction in HDRS scores = 5).
The pooled estimate of effect (SMD −0.08, 95% Grade 1, CI −0.45 to 0.30) was consistent with both a equivocal.
presence and absence of therapeutic benefit. Trialswere short term and the total sample size (141)was small.
Both Crocus and Fluoxetine groups showed a Grade 2, (Saffron) in mild to moderate depression.
significant improvement over the 6 weeks of inconclusive.
treatment (P b 0.0001). Crocus was found to havean effect similar to Fluoxetine (F = 0.13, df = 1,P = 0.71). Sample size was small (40).
Both Crocus and Imipramine groups showed Grade 2, significant improvement over the 6 weeks of inconclusive.
treatment (P b 0.0001). The difference betweenthe two was not significant as indicated by theeffect of group, the between-subjects factor(F = 2.91, df = 1, P = 0.09). Sample size was small(30).
More effective than placebo (P b 0.0001) with Grade 2, regard to improvement in the primary outcome inconclusive.
criterion HAMD21, which decreased from 12.59 (S.D. 1.39) to 6.81 (S.D. 3.75) in the neurapasgroup and from 12.47 (S.D. 1.40) to 12.79 (S.D.
2.46) in the placebo group. Sample size was small(67).
A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 Both the relaxation and CBT groups had Grade 2, significantly better BDI scores (P b 0.01) than inconclusive.
the tricyclic group post-treatment. 73% of therelaxation group improved to set BDI criteriapost-treatment compared to 82% for CBT and29% for tricyclics. The study had severalmethodological drawbacks, including smallsample size (37).
Nutritional ther apy—Docosahexaenoic acid Response (defined as N/=50% reduction in Grade 2, MADRS scores) rates were 27.8% in the DHA negative.
group and 23.5% in the placebo group, thedifference between which did not reach statisticalsignificance.
No evidence that tyrosine had antidepressant Grade 2, activity. Sample size small (65) and study negative.
duration short (4 weeks).
Less effective than Imipramine in mild to Grade 2, moderate depression ( F = 13.16, df = 1, negative.
P = 0.001), though a combination of Imipramineand Lavandula was more effective thanImipramine alone (F = 20.83, df = 1, P b 0.0001).
Sample size was small (45).
Participants in both music intervention groups Grade 3, showed significant improvements on all inconclusive.
measures at 8 weeks compared to wait listcontrols, with the improvements beingmaintained at 9 month follow-up (P b 0.05). Thesample was small (30), not representative of thegeneral population and double-blind conditionswere lacking.
Remission (HDRS =/b7) rates were 67%, Grade 3, compared to 93% in ECT group and 73% in negative.
tricyclic group, at the end of 4 weeks. Sample sizewas small (45) and double-blind conditionslacking.
5 of 12 patients with Major Depression showed Grade 3, reduction in depression scores on movement inconclusive.
group, uncontrolled therapy days compared to days without therapystudy (P b 0.05). The study had serious methodologicallimitations.
HAMD scores in the 5 patients with mild Grade 3, depression improved from 14.8 (S.D. 2.39) to 8.8 inconclusive.
(S.D. 3.63) (P = 0.039). The study had serious 19 cured, 16 improved, 5 failed, with total Grade 5, association between these substances and the potentially nine (SAMe) (), indicating that it may be of fatal Eosinophilia–Myalgia Syndrome has not been some benefit in Depression. The only systematic review and meta-analysis of the studies on SAMe, though notrecent (1994), showed a greater response rate when compared with placebo, with a global effect size rangingfrom 27% to 38% depending on the definition of We also found Grade 1 evidence for another response. The meta-analysis of 6 RCTs found that 70% Nutritional Therapy approach, with S-adenosyl methio- of subjects showed some response to SAMe, compared A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 with 30% for placebo. Furthermore, pooling of data from the 3 included), which evaluated the use of folate as 7 trials comparing SAMe with standard tricyclics found no difference indicating a comparable antidepressant 151 people assessed the use of folate in addition to other effect. However, most studies were methodologically treatment and found that adding folate reduced Hamilton flawed and brief in duration (b6 weeks). SAMe was well Depression Rating Scale scores on average by a further tolerated and caused few side-effects. Reliable compar- 2.65 points (95% confidence interval 0.38 to 4.93). One study involving 96 people assessed the use of folateinstead of the antidepressant trazodone and did not find a A Cochrane systematic review and meta-analysis of 7 RCTs on Acupuncture producedequivocal results. The results from 5 trials (409 par- A systematic review to determine the effectiveness of ticipants) included in the meta-analysis showed no inositol in treating depression identified four trials, with difference in the reduction in the severity of depression (HAM-D) compared to medication (WMD 0.53, 95% short-term trials of double-blind design. In one trial, CI −1.42 to 2.47). 4 trials (375 participants) reported on inositol was used as monotherapy. In the other studies it improvement in depression as an outcome (RR 1.20, was used in addition to conventional antidepressant 95% CI 0.94–1.51), again showing no differences agents. The pooled estimate of effect (SMD −0.08, 95% between groups. However, the evidence was insufficient CI −0.45 to 0.30) was consistent with both a presence to determine the efficacy of acupuncture vs. medication and absence of benefit. Thus, there was no clear evidence due to the poor methodological quality and reporting of these trials. There was insufficient data to demonstratewhether acupuncture is more effective than a wait-list control, non-specific or sham acupuncture control, orwhether acupuncture plus medication is more effective We found RCTs conforming to Grade 2 criteria as the highest level of available evidence on the effectivenessof Saffron (herbal medicine), homoeopathy, and relax- ation training in Depressive disorders, all of whichshowed inconclusive results. We also found RCTs eval- We found Grade 1 evidence pertaining to the effec- uating some nutritional therapy and herbal medicine approaches that showed conclusively negative results.
). However, this evidence is inconclusive. All 14 These were in addition to those RCTs evaluating CAM studies included in the systematic review had important approaches that had Grade 1 evidence.
methodological weaknesses. The participants in moststudies were community volunteers i.e. not clinical populations, and diagnosis depended on their score onthe Beck Depression Inventory. When compared with Two well-designed RCTs from the same research no treatment, exercise reduced symptoms of depression group compared the effectiveness of the herb Saffron (standardised mean difference in effect size −1.1 (95% (Crocus sativus L.) against standard antidepressants in confidence interval −1.5 to −0.6); weighted mean dif- mild to moderate depression, diagnosed according to the ference in Beck depression inventory −7.3 (−10.0 to Structured Clinical Interview for DSM-IV. In the first −4.6)). The effect size was significantly greater in those trials with shorter follow up and in two trials reported groups showed a significant improvement over the 6 only as conference abstracts. The effect of exercise was weeks of treatment (P b 0.0001). Crocus was found to similar to that of cognitive therapy (standardised mean have an effect similar to Fluoxetine (F = 0.13, df = 1, difference −0.3; 95% confidence interval −0.7 to 0.1).
P = 0.71). However, the sample size was small (40). In thesecond both Saffron and Imipramine groups showed a significant improvement inthe HAM-D over the 6 weeks of treatment (P b 0.0001).
A systematic review on the effectiveness of folate in The difference between the two was not significant as the treatment of depression found only one trial (out of indicated by the effect of group, the between-subjects A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 factor (F = 2.91, df = 1, P = 0.09). Sample size (30) was Lavandula tincture was more effective than Imipramine alone (F = 20.83, df = 1, P b.0001).
Another RCT studied the effectiveness of Neurapas We found studies conforming to Grade 3 criteria as balance, a complex homoeopathic remedy, which is a the highest level of available evidence on the effective- combination of St. John's wort, passion flower and ness of Aromatherapy massage, Movement Therapy, valerian extracts, in 67 adult patients with mild Music Therapy and Yoga in Depressive disorders. These depressive disorders according to ICD-10. The patients were in addition to those Grade 3 studies evaluating had a baseline Hamilton Rating Scale for Depression CAM approaches which had Grade 1 and 2 evidence.
It was not possible to draw firm conclusions from the results of these studies as all of them had significant was found to be more effective than placebo (P b 0.0001) with regard to improvement of the primary outcomecriterion HAMD21, which decreased from 12.59 (S.D.
1.39) to 6.81 (S.D. 3.75) in the neurapas group and from12.47 (S.D. 1.40) to 12.79 (S.D. 2.46) in the placebo An open RCT evaluated the effectiveness of Music Therapy in 30 older adults (aged 61–86 years) diagnosedwith major or minor depressive disorder ). They were randomly assigned to oneof three 8-week conditions: a home-based program Another RCT with a small sample (37) of patients where participants learned music listening stress reduc- with DSM-IV moderate depression, compared relaxation tion techniques at weekly home visits by a music training, cognitive-behavioural therapy and tricyclic therapist, a self-administered program where participants applied these same techniques with moderate therapist evidence in favour of relaxation training, with 73% of the intervention (a weekly telephone call), or a wait list relaxation group improving to set BDI criteria post- control. After 8 weeks, the Geriatric Depression Scale treatment, compared to 82% for CBT and 29% for scores of the 2 music groups were significantly better tricyclics. Both the relaxation and CBT groups had than those of the control group (P b 0.05). These significantly better BDI scores (P b 0.01) than the improvements were maintained over a 9-month follow- pharmacological treatment post-treatment. However, up period. The sample was small in size, not represen- the results need to be interpreted with caution because tative of the general population (only older adults, of the sample size, lack of control for the effects of predominantly female and highly educated) and double- attention from professionals (in the relaxation and CBT blind conditions were lacking. This result has not been groups) and reported non-compliance in the medication 5.4. Docosahexaenoic acid, tyrosine and Lavandulaangustifolia Mill Another open RCT compared the relative antidepres- sant efficacy of Sudarshan KriyaYoga (SKY) in RCTs conducted on the effectiveness of Docosahex- depression with electroconvulsive therapy (ECT) and aenoic acid (DHA, an omega-3 fatty acid) in major melancholic depressives (N = 45) were randomized naturally occurring aminoacid and neurotransmitter equally into 3 treatment groups and followed up over 4 weeks. Significant reductions in scores on the Beck both yielded unequivocally negative results. A small, Depression Inventory (BDI) and Hamilton Rating Scale preliminary RCT on the use of L. angustifolia Mill.
for Depression (HAM-D) occurred on successive Tincture, an herbal medication, in mild to moderate occasions in all 3 groups. At week 3, the SKY group had higher scores than the ECT group but was not preparation was unequivocally less effective than Imip- different from the Imiporamine group. Remission ramine. However, a combination of Imipramine and (HAM-D score of 7 or less) rates at the end of the trial A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 were 93, 73 and 67% in the ECT, Imipramine and SKY been little headway in using this evidence in providing groups, respectively. Due to the lack of double-blind conditions, small sample size and the short treatment This may be because the evidence for CAM effective- period, the data is insufficient to judge the value of SKY ness is viewed with scepticism. The sceptic may, however, point out that the available evidence is notconclusive enough. CAM treatments may be so different in terms of the individualisation of treatments, theintegrity of practitioner–patient relationships, the use of healing rituals and the subtlety and long time frames of with Major Depression were randomly assigned to outcomes expected, that they defy standard research movement therapy sessions on 7 of 14 days. Five of them showed a reduction in depression scores on movement therapy days compared to days without therapy (P b 0.05). The study had several methodological remains open to debate. Our review was intended to be drawbacks, including small sample size and lack of broad based, identifying the highest level of evidence available for each type of CAM in Depressive disorders.
We found systematic reviews and RCTs for 7 types of patients aged 31–59 years with a single episode of DSM CAM i.e. Herbal Medicine-St. John's wort (Hypericum IV Mild Depression received a 30-min aromatherapy perforatum), Nutritional Therapy-Tryptophan and 5- massage using essential oils of sweet orange, geranium hydroxy-tryptophan, S-adenosyl methionine (SAMe), and basil twice a week for 4 weeks. Patients were Folate and Inositol, Acupuncture and Exercise. Howev- submitted to the 17-item Hamilton Depression Rating er, all of these noted important methodological draw- Scale (HAM) and Profile of Mood States (POMS) backs in the included RCTs. Though the quality of trials 1 week before the first session and 1 week after the last was highly variable, most of them had small sample session. HAM score and the confusion–bewilderment sizes, short trial durations and poor monitoring of (C–B) score, one of the subscales of POMS, improved adverse events. In addition, many trials failed to describe from 14.8 ± 2.39 (Mean ± S.D.) to 8.8 ± 3.63 (P = 0.039) the methods used to achieve randomisation, or to and 62.2 ± 13.07 to 51.6 ± 8.05 (P = 0.043), respectively.
maintain allocation concealment. Successful blinding The study had several methodological drawbacks, was often questionable in trials of acupuncture and including small sample size, and the lack of blinding exercise, and to a lesser extent, herbal medicine. Data regarding dropouts and withdrawals was scarce, espe-cially in many of the older trials. Many older trials also lacked diagnostic precision. Intention-to-treat analyseswere rare. The systematic reviews were also constrained by the fact that few trials among the large number of approaches in Depressive disorders, they did not amount studies available were of sufficient quality to meet to the highest level of evidence available on those CAM inclusion criteria. Many of these findings corroborate approaches. Grade 4 studies were found only for those conclusions drawn by earlier researchers regarding the CAM approaches which had Grades 1, 2 and 3 evidence.
poor methodological quality of RCTs in CAM ( At Grade 5, we found a case series (n = 40) as the highest ). It is therefore clear that despite the avail- level of evidence on the use of Traditional Chinese ability of systematic reviews and RCTs, these types of CAM have not conclusively demonstrated efficacy over reported 19 remissions, 16 reductions in severity and 5 placebo, let alone standard antidepressants.
non-responders, with a total effective rate of 87.5%.
We found 7 RCTs evaluating the efficacy of 4 types of Though increasing numbers of trials are being CAM in depressive disorders. These included 4 RCTs undertaken in evaluating CAM approaches, there has which produced statistically positive results which A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 favoured the interventions–two on the Herbal medicine– differences between antidepressants and active placebos C. sativus L. (Saffron) comparing it against fluoxetine were small suggesting that unblinding effects may even and imipramine respectively, and one each on the inflate the efficacy of antidepressants in trials using inert Complex Homoeopathic remedy Neurapas balance (a combination of St. John's wort, passion flower and This also raises issues regarding the utility of the valerian extracts), and Relaxation Therapy. However, the hierarchy of evidence we used with respect to CAM trials.
trials shared many of the methodological drawbacks Where should a CAM trial, which was adequately discussed in the previous section. All had uniformly randomized and controlled, but could not be blinded for small sample sizes. The Relaxation Therapy trial relied the very reasons discussed earlier, be placed in the exclusively on subjective scales to measure outcomes.
hierarchy of evidence, with its strict definitions of what Successful blinding was questionable in the trials of constitutes each level? Here, the rating scale used as a homoeopathic medicine and relaxation therapy, and to a measure of effectiveness in the National Service Frame- lesser extent, herbal medicine. Information on power work for Mental Health in the United Kingdom ; calculations were either insufficient for a critical appraisal or absent. Adequate reporting of adverse which draws on a ‘synthesis of evidence’ from research effects was restricted to the herbal medicine trials. In findings, and rated on a five-point ordinal scale according addition, none of the findings have been replicated in to their inferential power, appears to do better. This is other trials or settings. It must be noted that the two RCTs related to the definitions of the respective levels of on the Herbal medicine Saffron (which were from the evidence being more pragmatic and hence, oriented to same research group) had methodologically robust everyday clinical reality. The inclusion of the opinions of designs, but were hampered by their sample sizes.
service users and carers under the umbrella of expert The 3 remaining RCTs included the Nutritional opinion, as Type V evidence, is representative of this and interventions—Docosahexaenoic acid (DHA) and Ty- of relevance to CAM. However, the comparative rosine, and the Herbal medicine L. angustifolia Mill.
strengths and utilities of each hierarchy are open to debate.
Tincture, all of which, though sharing many of the samemethodological drawbacks, produced statistically nega- tive results with respect to their use as monotherapy indepression.
As mentioned earlier, the Grade 4 studies we found did not amount to the highest level of evidence available on those CAM approaches. At Grade 5, we found a caseseries of 40 cases on the use of Traditional Chinese We found studies on 4 types of CAM-Music Therapy, Medicine in melancholia. This was the highest level of Yoga (Sudarshan Kriya), Movement Therapy and evidence available on the use of this particular CAM Aromatherapy massage. All the studies had serious approach and was based on clinical impressions of what methodological limitations. All had small sample sizes constituted remission and response. It also used an and inadequate information on power calculations. All approach to the diagnosis of melancholia based on the trials were open, as the nature of these interventions traditions of Chinese medicine. Despite the obvious probably rendered blinding difficult. The issues raised by contradiction this presented to our stated focus on ICD/ DSM depressive disorders, the evidence was included in application of standard research methods to CAM are the review as it was representative of a large body of case germane in this context. Blinding is difficult when anintervention such as Yoga is to be compared against a ‘placebo’ intervention, let alone an antidepressant. Such Rating scale used as a measure of effectiveness in the national service interventions can, in theory, be compared against similar interventions of the sham variety (the ‘placebo’ inter- vention) as has been done often in trials of acupuncture.
including at least one randomised controlled trial.
However, in order to be a genuine placebo, the control At least one good randomised, controlled trial.
Type III evidence At least one well-designed intervention procedure must be convincing, visible and should mimic, in all respects, apart from a physiological effect, Type IV evidence At least one well-designed observational study.
the real active treatment. This has been found to be Expert opinion, including the opinion of service A Cochrane systematic review has found that A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 series on the use of Traditional Chinese Medicine in to specify (). In response to the methodo- logical difficulties that have arisen while evaluating such Despite the potential drawbacks of its standard RCT- interventions, the Medical Research Council (MRC) has based research methods with respect to CAM, the developed a framework to evaluate these complex Cochrane Collaboration has facilitated CAM research.
interventions. It has been emphasized that this frame- By March 2004, the Cochrane Collaboration had 145 work should not be read as an inflexible ‘to do list’, but completed reviews of randomised controlled trials of rather as advice to apply to the extent to which it is complementary and alternative therapies: a third showed relevant at a given stage of the project (In a positive or possibly positive effect, although over half effect, this framework offers an approach to designing found insufficient evidence to make such judgments studies in keeping with the current levels of knowledge in the field, and encourages pragmatic and exploratory evaluation remains the double blind, randomized research designs as appropriate, an approach that appears particularly relevant to designing studies in CAM. Other It is clear from our review that despite the availability responses to the difficulties encountered in CAM of systematic reviews in 7 complementary therapies and research have included pragmatic trial designs, nested RCTs in 12, these types of CAM have not conclusively qualitative studies, and the use of real world observa- demonstrated efficacy over placebo in depression. Some tional data to create an “evidence house” ).
of the methodological problems in applying the RCT As discussed earlier, the existing systematic reviews model to complementary therapies may partly explain have been constrained by the fact that few trials among the large number of studies available were of sufficient The application of the RCT model to many types of quality to meet inclusion criteria. Such pragmatic CAM is potentially fraught with difficulties. RCTs are approaches may address that particular problem. How- specifically designed to cancel out the human factors (in ever, the question remains whether complementary particular the non-specific effects of the therapeutic therapies will be shown to have effects of sufficient relationship) and confounding variables (e.g. regression magnitude when larger trials with robust designs are to the mean and patient variability). Many complementary therapies on the other hand intentionally involve non- And then again, the growth in evidence on the clinical specific elements, and depend on factors like individual effectiveness of some complementary therapies in other resilience and changing patients' beliefs and behaviour.
conditions has not been matched by evidence of cost Obviously CAM is quite different from simply taking a effectiveness. Complementary treatments present an pill. This makes their evaluation difficult though the additional healthcare cost in four out of the five rigorous difficulties are not insurmountable by pragmatic trial cost effectiveness studies conducted in the UK most CAM use occurs among patients with chronic acupuncture and spinal manipulation, they do raise an diseases and relapsing dysfunction e.g. chronic pain, issue that CAM researchers in mental health will have to anxiety, chronic fatigue and chronic musculoskeletal address soon, as health services find themselves with increasingly finite economic resources.
often unsatisfactory or relatively ineffective. In these areasit is difficult to establish clearly defined end-points by which to determine objective outcomes; they are alsomore difficult to evaluate using the conventional This review was limited by the fact that the searches randomised controlled trial. Therefore the outcomes were restricted to the English language. It may be that may be broader and less predictable than those standard significant research exists, outside the ambit of English language journals and western research models, of which However, many of these methodological problems we are unaware. Our focus on the use of CAM as are shared by complex interventions for chronic monotherapy excluded studies that investigated the use of CAM as an adjunct to conventional treatment. We also ). Complex interventions are interventions, whether excluded studies in which the primary therapeutic curative or preventative, comprising a number of intervention under evaluation consisted of a combination separate elements, which seem essential to the proper of two or more complementary therapies. Research does functioning of the intervention although the ‘active exist that addresses these particular areas and it is ingredient’ of the intervention that is effective, is difficult possible that some of this evidence is relevant to the A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 treatment of depressive disorders. Despite our best Campbell, M., Fitzpatrick, R., Haines, A., Kinmonth, A.L., Sander- efforts to ensure the inclusion of all CAM approaches, cock, P., Spiegelhalter, D., et al., 2000. Framework for design andevaluation of complex interventions to improve health. BMJ 321, given the rich diversity of complementary therapies, it is possible that the list of CAM approaches we arrived at is Canter, P.H., Coon, J.T., Ernst, E., 2005. Cost effectiveness of not a comprehensive one. This again raises the complementary treatments in the United Kingdom: systematic possibility that we may have missed important research.
Cooper, B., 2003. Evidence-based mental health policy: a critical appraisal. Br. J. Psychiatry 183, 105–113.
Davidson, J.R.T., Meltzer-Brody, S.E., 1999. The under-recognition and under-treatment of depression: what is the breadth and depth of 19 studies that used CAM as monotherapy in depression the problem? J. Clin. Psychiatry 60 (suppl 7), 4–9.
fulfilled the criteria set for inclusion in the final review.
Department of Health, 1999. National Service Framework for Mental Systematic reviews formed the highest level of evidence in Health: Modern Standards and Service Models. Department ofHealth, London.
7 complementary therapies and RCTs in 5. However, Ernst, E., et al., 2001. The Desktop Guide to Complementary and neither these nor those complementary therapies with Alternative Medicine—An Evidence-Based Approach. Mosby.
lower levels of evidence have conclusively demonstrated Ernst, E., White, A.R., 2000. The BBC survey of complementary efficacy over placebo in depression. All studies had medicine use in the UK. Complement. Ther. Med. 8, 32–36.
significant methodological problems. The widely Fabrega, H., 2002. Medical validity in Eastern and Western traditions.
referenced hierarchy of evidence we used has limited Gelenberg, A.J., Wojcik, J.D., Falk, W.E., Baldessarini, R.J., 1990.
utility in grading trials of CAM. Despite the drawbacks of Tyrosine for depression: a double-blind trial. J. Affect. Disord. 19 its RCT based research model with respect to CAM, the Cochrane collaboration has facilitated CAM research. The Government Response to The House of Lords Select Committee on RCT model and the principles underlying many types of Science and Technology's Report on Complementary andAlternative Medicine. Presented to Parliament by the Secretary CAM are dissonant, making its application in the of State for Health By command of Her Majesty. Department of evaluation of those types of CAM difficult. These Health 2001, March. Crown Copyright 2001.
methodological problems are similar to those encountered Gray Muir, J.A., 1997. Evidence based Healthcare: how to make health in the evaluation of complex interventions for chronic policy and management decisions. Churchill Livingstone, New conditions in conventional health care. The MRC Hanser, S.B., Thompson, L.W., 1994. Effects of music therapy strategy framework for the evaluation of such interventions can on depressed older adults. J. Gerontol. 49, 265–269.
be used in CAM, as can pragmatic trial designs, nested House of Lords Select Committee on Science and Technology qualitative studies and observational data. Unless CAM Complementary and Alternative Medicine, 2000. 6th Report approaches are shown to have effects of sufficient 1999–2000. Stationery Office, London.
magnitude in larger trials with robust designs, their utility Janakiramiah, N., Gangadhar, B.N., Naga Venkatesha Murthy, P.J., Harish, M.G., Subbakrishna, D.K., Vedamurthachar, A., 2000.
in the treatment of depression remains limited. This Antidepressant efficacy of Sudarshan Kriya Yoga (SKY) in information is of relevance to service users, planners, melancholia: a randomised comparison with electroconvulsive general practitioners and mental health professionals, given therapy (ECT) and imipramine. J. Affect. Disord. 57 (1–3), 255–259.
the increasing use of CAM by those who are depressed.
Jonas, W.B., 2001. The evidence house: how to build an inclusive base for complementary medicine. West. J. Med. 175, 79–80.
Kendrick, T., 2000. Why can't GPs follow guidelines on depression? Kessler, R.C., Soukup, J., Davis, R.B., Foster, D.F., Wilkey, S.A., Van Akhondzadeh, S., Kashani, L., Fotouhi, A., Jarvandi, S., Mobaseri, Rompay, M.I., Eisenberg, D.M., 2001. The use of complementary M., Moin, M., Khani, M., Jamshidi, A.H., Baghalian, K., and alternative therapies to treat anxiety and depression in the Taghizadeh, M., 2003. Comparison of Lavandula angustifolia United States. Am. J. Psychiatry 158 (2), 289–294.
Mill. Tincture and Imipramine in the treatment of mild to moderate Lawlor, D.A., Hopker, S.W., 2001. The effectiveness of exercise as an depression: a double-blind, randomized trial. Prog. Neuro- intervention in the management of depression: systematic review psychopharmacol. Biol. Psychiatry 27 (1), 123–127.
and meta-regression analysis of randomised controlled trials. BMJ Akhondzadeh, S., Fallah-Pour, H., Afkham, K., Jamshidi, A.H., Khalighi- Cigaroudi, F., 2004. Comparison of Crocus sativus L. and imipramine Linde, K., Jonas, W.B., Melchart, D., Willich, S., 2001. The methodo- in the treatment of mild to moderate depression: a pilot double-blind logical quality of randomized controlled trials of homeopathy, herbal randomized trial. BMC Complement. Altern. Med. 4, 12.
medicines and acupuncture. Int. J. Epidemiol. 30, 526–531.
Astin, J.A., 1998. Why patients use alternative medicine: results of a Linde K, Mulrow C.D., Berner, M., Egger, M. St John's Wort for national study. JAMA 279, 1548–1553.
depression. The Cochrane Database of Systematic Reviews 2005, Issue Bressa, G.M., 1994. S-adenosyl-L-methionine (SAMe) as antidepres- 3. Art. No.: CD000448. .CD000448.pub2.
sant: meta-analysis of clinical studies. Acta Neurol. Scand., Suppl.
MacLennan, A.H., Wilson, D.H., Taylor, A.W., 1996. Prevalence and cost of alternative medicine in Australia. Lancet 347, 569–573.
A.F. Thachil et al. / Journal of Affective Disorders 97 (2007) 23–35 Manheimer, E., Berman, B., Dubnick, H., Beckner, W., 2004. Cochrane Taylor, M.J., Carney, S., Geddes, J., Goodwin, G. Folate for depressive Reviews of Complementary and Alternative Therapies: Evaluating disorders. The Cochrane Database of Systematic Reviews 2003, Issue Taylor, M.J., Wilder, H., Bhagwagar, Z., Geddes, J. Inositol for Marangell, L.B., Martinez, J.M., Zboyan, H.A., Kertz, B., Kim, H.F., depressive disorders. The Cochrane Database of Systematic Puryear, L.J., 2003. A double blind, placebo-controlled study of Reviews 2004, Issue 1. Art. No.: CD004049. the omega-3 fatty acid docosahexaenoic acid in the treatment of major depression. Am. J. Psychiatry 160 (5), 996–998.
The World Health Report, 2001. Mental Health: new understanding, Micozzi, M.S., 2005. Fundamentals of Complementary and Alterna- new hope. World Health Organization, Geneva.
tive Medicine, 3rd edition. Elsevier, New York.
Thomas, K.J., Nicholl, J.P., Coleman, P., 2001. Use and expenditure on Moncrieff, J., Wessely, S., Hardy, R. Active placebos versus antidepres- complementary medicine in England: a population based survey.
sants for depression. The Cochrane Database of Systematic Reviews 2004, Issue 1. Art. No.: CD003012. .
Thompson, T., Feder, G., 2005. Complementary therapies and the MRC, 2000. A framework for development and evaluation of RCTs Urlea-Schön, I., Wartenberg-Demand, A., McGregor, G., 2003.
for complex interventions to improve health.
Efficacy of a triple herbal preparation in mild depressive disorders: Murphy, G.E., Carney, R.M., Knesevich, M.A., Wetzel, R.D., results of a randomised placebo-controlled trial. Focus Altern.
Whitworth, P., 1995. CBT, relaxation training and TCA medication in the treatment of depression. Psychol. Rep. 77, 403–420.
Vickers, A.J., 2001. Message to Complementary and Alternative Noorbala, A.A., Akhondzadeh, S., Tahmacebi-Pour, N., Jamshidi, A.H., Medicine: evidence is a better friend than power. BMC Comple- 2005. Hydro-alcoholic extract of Crocus sativus L. versus fluoxetine in the treatment of mild to moderate depression: a double-blind, Vickers, A., Cassileth, B., Ernst, E., Fisher, P., Goldman, P., Jonas, W., randomized pilot trial. J. Ethnopharmacol. 97 (2), 281–284.
Kang, S.K., Lewith, G., Schulz, K., Silagy, C., 1997. How should Okamoto, A., Kuriyama, H., Watanabe, S., Aihara, Y., Tadai, T., we research unconventional therapies? A panel report from the Imanishi, J., Fukui, K., 2005. The effect of aromatherapy massage on Conference on Complementary and Alternative Medicine Research mild depression: a pilot study. Psychiatry Clin. Neurosci. 59 (3), 363.
Methodology. National Institutes of Health. Int. J. Technol. Assess.
Raben, H., Aggernaes, K.H., 1991. Psykiatriske patienters brug af alternativ behandling. Ugeskr. Laeger 153, 782–784.
Walach, H., 2001. The efficacy paradox in randomized controlled trials Razali, S.M., Najib, M.A., 2000. Help seeking pathways among Malay of CAM and elsewhere: beware of the placebo trap. J. Altern.
psychiatric patients. Int. J. Soc. Psychiatry 46 (4), 281–289.
Scott, J., Dickey, B., 2003. Global burden of depression: the White, P., Lewith, G., Hopwood, V., Prescott, P., 2003. The ‘placebo’ intersection of culture and medicine. Br. J. Psychiatry 183, 92–94.
needle, is it a valid and convincing control for use in acupuncture Shaw, K., Turner, J., Del Mar, C. Tryptophan and 5-Hydroxytrypto- trials?—A randomised, single blind, crossover pilot trial. Focus phan for depression. The Cochrane Database of Systematic Altern. Complement. Ther. 8, 554–555.
Reviews 2002, Issue 1. Art. No.: CD003198. Wu, J., An, H., Li, Y., Duan, L., 1999. TCM treatment with the modified Wendan Tang in 40 cases of melancholia. J. Tradit. Chin. Med. 19 (4), Smith, C.A., Hay, P.P.J. Acupuncture for depression. The Cochrane Database of Systematic Reviews 2004, Issue 3. Art. No.:CD004046. .CD004046.pub2.
Stewart, N.J., McMullen, L.M., Rubin, L.D., 1994. Movement therapy with depressed inpatients: a randomized multiple single case-design. Arch. Psychiatr. Nurs. 8, 22–29.

Source: http://www.arttherapyaustralia.com.au/wp-content/uploads/2013/06/evidence-base-of-complementary-therapies-in-depression.pdf