Journal of Psychiatric Research 57 (2014) 165e175

Contents lists available at ScienceDirect

Journal of Psychiatric Research journal homepage: www.elsevier.com/locate/psychires

A systematic review on the efficacy, safety and types of Chinese herbal medicine for depression Wing-Fai Yeung a, Ka-Fai Chung b, *, Ka-Yan Ng b, Yee-Man Yu b, Eric Tat-Chi Ziea c, Bacon Fung-Leung Ng c a b c

School of Chinese Medicine, University of Hong Kong, Hong Kong, China Department of Psychiatry, University of Hong Kong, Hong Kong, China The Chinese Medicine Department, Hospital Authority, Hong Kong, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 March 2014 Received in revised form 5 May 2014 Accepted 28 May 2014

Chinese herbal medicine (CHM) is one of the commonly used complementary and alternative medicine therapies for major depressive disorder. The objective of this study was to review the efficacy, safety and types of CHM for depression. We systematically searched key databases (9 Chinese and 7 English) up until May 2013 for randomized controlled trials (RCTs) and examined 7 systematic reviews for additional articles. Methodological quality was assessed by modified Jadad scale and Cochrane's risk of bias assessment. Only studies with moderate methodological quality, defined as modified Jadad scale score
3, were included in meta-analysis for efficacy. Of the 296 RCTs that were assessed in details, 278 (93.9%) had modified Jadad scale score < 3, and only 21 scored
3. The frequently used formulas were Xiao Yao decoction, Chaihu Shugan decoction and Ganmai Dazao decoction; while Chaihu, Bai Shao and Fu Ling were the frequently used single herb. Meta-analyses showed that CHM monotherapy was better than placebo and as effective as antidepressants in reducing Hamilton Depression Rating Scale (HDRS) score (CHM vs. placebo: mean difference: 7.97, 95% CI: 10.25 to 5.70, P < 0.00001, 2 studies; CHM vs. antidepressants: mean difference: 0.01, 95% CI: 0.28 to 0.30, P ¼ 0.95, 7 studies). CHM were associated with less adverse events than antidepressants, and adding CHM to antidepressants reduced adverse events. Despite the overall positive results, due to the small number of studies with sufficient methodological quality, it is premature to accurately conclude the benefits and risks of CHM for depression. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Chinese herbal medicine TCM Depression Systematic review Meta-analysis

1. Objectives of the study and background According to the Global Burden of Disease Study 2010, major depressive disorder (MDD) was ranked the second leading cause of years lived with disability, after low back pain, accounting for 8.2% of all years lived with disability (Ferrari et al., 2013). The World Mental Health Survey Initiative showed that the average lifetime prevalence for major depressive episode based on the Diagnostic and Statistical Manual, Fourth Edition (DSM-IV, American Psychiatric Association, 1994) was 14.6% in 10 high-income countries and 11.1% in 8 low- to middle-income countries (Bromet et al., 2011). MDD not only affects individuals' work, school and daily life, but it also affects their life satisfaction and perceived well-being.

* Corresponding author. Department of Psychiatry, University of Hong Kong, Pokfulam Road, Hong Kong, China. Tel.: þ86 852 22554487; fax: þ86 852 28551345. E-mail address: [email protected] (K.-F. Chung). http://dx.doi.org/10.1016/j.jpsychires.2014.05.016 0022-3956/© 2014 Elsevier Ltd. All rights reserved.

As the number of people with depression is rapidly increasing across the world (Baxter et al., 2014), more effective treatments should be identified in order to reduce the potential harms MDD brings to sufferers' lives. Pharmacotherapy is currently the most commonly used treatment for MDD because of its reported effectiveness. However, complaints such us nausea, headache, insomnia, agitation, weight gain daytime somnolence and sexual dysfunction are often reported during the course of treatment, leading to treatment termination in some patients. Psychological treatments for depression are also commonly used. Despite its proven effectiveness, the use of psychotherapy is limited by its time-intensive nature, limited access to skilled providers, high cost, and requirement of patients' participation and motivation. According to the World Health Organization Mental Health Atlas 2011, psychosocial interventions were not readily available in more than half of the countries surveyed, especially the low income countries (World Health Organization, 2011). Faced with the limitations of the

166

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

currently available treatments, the use of complementary and alternative medicine for depression is common. A national representative survey in the United States found that 53.6% of people with self-reported depression reported using some forms of complementary and alternative therapies to treat depression during the past 12 months (Kessler et al., 2001). A large number of studies have been performed to examine the effectiveness of complementary and alternative therapies for mood disorders (Qureshi and AIBedah, 2013), suggesting that there is a demand for treatments other than pharmacotherapy and psychotherapy. Chinese herbal medicine is one of the most commonly used modalities of complementary and alternative medicine therapies, especially in Chinese culture (Hsu et al., 2008). There have been previous systematic reviews on specific CHM formulas, including Chaihu-Shugan-San (Wang et al., 2012) and Xiao-Yao-San (Qin et al., 2011; Zhang et al., 2012). These reviews were limited by the poor methodological quality of the studies included in analysis and the grouping of bipolar disorder as depression. There were two other systematic reviews on CHM for depression (Butler and Pilkington, 2013; Zhao et al., 2009). The review by Butler and Pilkington (2013) was based on previous systematic reviews supplemented by an update search of English databases, whereas the study by Zhao et al. (2009) searched only one Chinese database. In addition, the pattern of CHM use for depression was not examined in previous reviews. In view of the shortcomings of previous studies, this systematic review aimed to: (1) summarize the efficacy and safety of CHM, as either monotherapy or adjunct therapy, in treating depression, with attention to studies with better methodological quality and (2) determine the most commonly used CHM formulas and single herbs for the treatment of depression. 2. Materials and methods Nine Chinese language databases (China Journals Full-text Database, China Proceedings of Conference Full-text Database, Chinese Biomedical Literature Database, China Doctor Dissertations Full-text Database, China Master Theses Full-text Database, Chinese Science and Technology Documents Database, Chinese Dissertation Document Bibliography Database, Taiwan Electronic Periodical Services, and WanFang Database) and seven English language databases (MEDLINE, EMBASE, Cochrane Central Register of Controlled Trials, Cumulative Index to Nursing and Allied Health Literature, Allied and Complementary Medicine, PsycINFO and ProQuest Dissertations and Theses A&I) were searched by two researchers independently up until May 2013 using the grouped terms (depression* OR depressive* OR dysthymia* OR mood disorder* OR “affective disorder*” OR “affective symptoms” OR MDD) AND (Chinese herb* OR herbal medicine* OR traditional Chinese medicine* OR TCM OR Chai-Hu-Shu-Gan-San OR ChaiHuShuGan* OR Xiao-Yao-San OR Xiao Yao* OR Ban-Xia-Hou-Pu-Tang OR Ban Xia Hou Pu* OR Gan-Mai-Da-Zao-Tang OR GanMai DaZao* OR Gui-PiTang OR GuiPi* OR Wen-Dan-Tang OR WenDan OR Yue-Ju-Wan OR Yue-Ju) and their equivalent Chinese terms. The reference lists of the included papers and previous systematic reviews (Kou and Chen, 2012; Wang et al., 2012; Qin et al., 2010; Qin et al., 2011; Butler and Pilkington, 2013; Zhao et al., 2009; Zhang et al., 2012) were further searched for relevant articles. We included saffron as a CHM because saffron has a long history of use in China despite its Persian origin. The use of saffron was documented in “Compendium of Materia Medica”, one of the most respected Chinese medical texts written in 1578, which termed saffron as “foreign red flower” (Feng, 2014). The herb is now named as “Western red flower” in the pharmacopoeia of China (Ministry of Health of the People's Republic of China (2010)). There was no language restriction in our search.

Studies included in this review were randomized or quasirandomized clinical trials that examined participants with depression according to one of the following diagnostic criteria: the DSM-IV (American Psychiatric Association, 1994), Research Diagnostic Criteria (RDC, Spitzer et al., 1989), Chinese Classification of Mental Disorders, Second-Revised/Third Edition (CCMD-2-R/3, Chinese Psychiatric Society, 2001) or other relevant criteria. The included studies should include one of the following comparisons: (1) CHM vs. placebo, antidepressants, psychotherapy, or routine care; (2) CHM in combination with antidepressants, psychotherapy, or routine care vs. antidepressants, psychotherapy, or routine care alone; and (3) CHM in combination with placebo antidepressants vs. antidepressants in combination with CHM placebo. In addition, at least one of the following outcome measures was used: (1) selfrating scales, (2) clinician-rated scales, or (3) effective rate, which was often defined as the proportion of subjects who had at least 30% reduction in depression score. Secondary outcome examined in this study was the occurrence of adverse events. Two authors (KN and YY) searched the databases and selected the relevant publications independently. Any disagreement about the eligibility of a study was resolved by discussion, and consultation with the senior authors (WY and KC). One author extracted the data (KN) and the other (YY) checked the extracted data. For each study, we examined the study design, patients' characteristics including age, gender, and duration of depression, CHM treatment, control intervention and outcome parameter. We analyzed the methodological quality of the studies using the modified Jadad scale (Jadad et al., 1996; White and Ernst, 1999) and the Cochrane's risk of bias assessment (Higgins and Green, 2011). The modified Jadad scale score ranges from 1 to 5; points are awarded if study: is described as randomized, 1 point; has appropriate randomization method, 1 point; is described as subject-blinded, 1 point; is described as evaluator-blinded, 1 point; and has description of withdrawals and dropouts, 1 point. Studies with a modified Jadad score
3 were considered to be moderate-quality randomized controlled trials (RCTs) (Manchikanti et al., 2011). The risk of bias assessment appraises a study in six domains: adequate sequence generation, allocation concealment, blinding of participants, personnel and outcome assessors, incomplete outcome data, selective outcome reporting and other sources of bias. Each domain can be rated as “yes” (low risk of bias), “no” (high risk of bias), or “unclear” (uncertain risk). A metaanalysis for efficacy would be performed only if studies were similar in clinical characteristics and had moderate methodological quality (modified Jadad score
3). Data were summarized using risk ratio (RR) with 95% confidence intervals (CI) for binary outcome; mean difference (MD) or standardized mean difference (SMD), if different studies measured the same outcome in different scales, with 95% CI was used for continuous outcome. The incidence of any adverse events was summarized using rate ratio (RaR). 3. Results 3.1. Description of the paper selection process and overview of the reviewed studies The search yielded 5097 potential titles, of which 929 were duplicate records and 3594 were excluded for reasons of irrelevance. The full text of 574 articles were retrieved for assessment, of which 278 were excluded for various reasons (Fig. 1). Full details of the excluded studies are available from the authors upon request. Of the remaining 296 studies that were examined in details, 104 studies used Western antidepressants as comparators, 108 studies on adjunctive CHM with antidepressants vs. antidepressants alone, while only seven studies used placebo control. The sample size of the 296 studies ranged from 30 to 1024, with a total of 24,876

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

167

Fig. 1. Flowchart of this systematic review.

subjects. The studies were carried out in China or Iran and were published in either Chinese or English language journals. Onehundred and ninety-four (65.5%) of the 296 studies found CHM was significantly more effective than control interventions for the treatment of depression (Table 1). 3.2. Description of the CHM formulas Three-hundred forty-five CHM formulas were tested in the 296 reviewed studies; 223 were standardized formulas while the other

122 were prescribed based on individuals' characteristics and TCM patterns. Table 2 presents the 10 most frequently used CHM formulas, their basic compositions, and the 10 most frequently used single herbs. The most frequently used formula was Xiao Yao decoction or its modification, which were examined 90 times, accounting for 26.1% of the total number of formulas tested, followed by Chaihu Shugan decoction (10.4%), and Ganmai Dazao decoction (4.1%). With regard to single herb, Chaihu (Bupleurum chinense DC.) was the most frequently used single herb for the treatment of depression and it was one of the herbal components in 183 of the

Table 1 Study design and methodological quality of the 296 included studies. Intervention

CHM vs. AD CHM vs. placebo CHM vs. PT CHN vs. RC CHM þ PT vs. AD þ PT CHM þ RC vs. AD þ RC CHM þ PT þ RT vs. PT þ RC CHM þ AD vs. AD CHM þ placebo vs. AD þ placebo CHM þ AD vs. placebo þ AD CHM þ AD þ PT vs. AD þ PT CHM þ AD þ RC vs. AD þ RC 3-arm design Total (%)

No. of studies

No. of studies showing CHM more effective

No. of randomized subjects

Total modified Jadad score I

II

III

IV

V

104 6 3 8 7 19 1 108 1 2 3 15 19 296 (100)

47 6 3 7 5 11 1 84 1 2 2 12 13 194 (64.9)

9319 488 535 602 404 1504 85 8465 66 418 213 1352 1425 24,876

66 0 3 7 6 11 0 72 0 0 3 12 10 190 (64.2)

30 1 0 1 1 6 1 35 0 0 0 3 7 85 (28.7)

5 0 0 0 0 2 0 1 0 0 0 0 2 10 (3.4)

0 3 0 0 0 0 0 0 1 2 0 0 0 6 (2.0)

3 2 0 0 0 0 0 0 0 0 0 0 0 5 (1.7)

AD, antidepressants; CHM, Chinese herbal medicine; PT, psychotherapy; RC, routine care.

No. of studies with modified Jadad score
3 8 5 0 0 0 2 0 1 1 2 0 0 2 21 (7.1)

168

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

Table 2 The 10 most frequently used Chinese herbal formulas and single herbs for depression (Total number of formulas ¼ 345). Frequently used Chinese herbal formula

Content

Xiaoyao decoction or its modification

90 (26.1%) Dang Gui (Angelica sinensis (Oliv.) Diels); Fu Ling (Poria Cocos (Schw) Wolf.); Zhi Zi (Gardenia jasminoides J.Ellis); Bo He (Mentha haplocalyx Briq..); Bai Shao (Paeonia lactiflora Pall.); Chai Hu (Bupleurum chinense DC.); Gan Cao (Glycyrrhiza uralensis Fisch.); Bai Zhu (Atractylodes macrocephala Koidz.); Mu Dan Pi (Paeonia  suffruticosa Andrews); Jiang (Zingiber officinale Roscoe) 36 (10.4%) Xiang Fu (Cyperus rotundus L); Bai Shao (Paeonia lactiflora Pall.); Chai Hu (Bupleurum chinense DC.); Chuan Xiong (Ligusticum striatum DC.); Zhi Ke (Citrus  aurantium L.); Chen Pi (Citrus reticulata Blanco); Gan Cao (Glycyrrhiza uralensis Fisch.) Gan Cao (Glycyrrhiza uralensis Fisch.); Xiao Mai (Triticum aestivum L.); Da Zao 14 (4.1%) (Ziziphus jujube Mill.) Ren Shen (Panax ginseng C.A.Mey.); Huang Qi (Astragalus membranaceus (Fisch.) 10 (2.9%) Bunge); Bai Zhu (Atractylodes macrocephala Koidz.); Fu Ling (Poria Cocos (Schw) Wolf.); Dang Gui (Angelica sinensis (Oliv.) Diels); Zao Ren (Ziziphus jujuba var. spinosa (Bunge) Hu ex H.F.Chow); Gui Yuan Rou (Dimocarpus longan Lour.); Yuan Zhi (Polygala tenuifolia Willd.); Mu Xiang (Saussurea costus (Falc.) Lipsch.); Gan Cao (Glycyrrhiza uralensis Fisch.); Jiang (Zingiber officinale Roscoe); Da Zao (Ziziphus jujube Mill.) 5 (1.4%) Ban Xia (Pinellia ternata (Thunb.) Makino); Ju Hong (Citrus reticulata Blanco); Fu Ling (Poria Cocos (Schw) Wolf.); Gan Cao (Glycyrrhiza uralensis Fisch.); Zhu Ru (Arundarbor angulata (Munro) Kuntze); Zhi Ke (Citrus  aurantium L.); Jiang (Zingiber officinale Roscoe); Da Zao (Ziziphus jujube Mill.) Ban Xia (Pinellia ternata (Thunb.) Makino); Hou Pu (Magnolia officinalis Rehder & E.H.Wilson); Fu Ling (Poria Cocos (Schw) Wolf.); Jiang (Zingiber officinale Roscoe); Gan Su Ye (Perilla frutescens (L.) Britt) 5 (1.4%) Chai Hu (Bupleurum chinense DC.); Long Gu (Fossilia Ossis Mastodi); Huang Qin (Scutellaria baicalensis Georgi); Jiang (Zingiber officinale Roscoe); Qian Dan (Lead tetroxide); Ren Shen (Panax ginseng C.A.Mey.); Gui Zhi (Cinnamomum cassia (L.) J.Presl); Fu Ling (Poria Cocos (Schw) Wolf.); Ban Xia (Pinellia ternata (Thunb.) Makino); Da Huang (Rheum palmatum L.); Mu Li (Ostrea gigas Thunb); Da Zao (Ziziphus jujube Mill.) Chai Hu (Bupleurum chinense DC.); Huang Qin (Scutellaria baicalensis Georgi); Yu 5 (1.4%) Jin (Curcuma wenyujin Y.H.Chen & C.Ling); Jiang (Zingiber officinale Roscoe); Bai He (Lilium brownii F.E.Br. ex Miellez); Mai Men Dong (Liriope spicata Lour.); Xi Yang Shen (Panax quinquefolius L.) 5 (1.4%) Huang Qi (Astragalus membranaceus (Fisch.) Bunge); Shan Yao (Dioscorea oppositifolia L.); Fu Ling (Poria Cocos (Schw) Wolf.); Ji Nei Jin (Gallus gallus domesticus Brisson); Sha Ren (Amomum villosum Lour.); Wu Mei (Prunus mume (Siebold) Siebold & Zucc.); Bin Lang (Areca cathecu L.) 5 (1.4%) Ren Shen (Panax ginseng C.A.Mey.); Lu Rong (Cervus nippon Temminck); Sheng Di Huang (Rehmannia glutinosa (Gaertn.) DC.); Shan Zhu Yu (Cornus officinalis Siebold & Zucc.); Nu Zhen Zi (Ligustrum lucidum W.T.Aiton); Shui Hong Hua Zi (Polygonum Orientalis (L.) Spach) Chai Hu (Bupleurum chinense DC.) (53.0%), Bai Shao (Paeonia lactiflora Pall.) (42.3%), Fu Ling (Poria Cocos (Schw) Wolf.) (41.7%), Yu Jing (Curcuma wenyujin Y.H.Chen & C.Ling) (40.0%), Dang Gui (Angelica sinensis (Oliv.) Diels) (36.2%), Bai Zhu (Atractylodes macrocephala Koidz.) (31.6%), Gan Cao (Glycyrrhiza uralensis Fisch.) (30.1%), Chuan Xiong (Ligusticum striatum DC.) (25.8%), Shi Chang Pu (Acorus tatarinowii Schott) (25. 2%), Xiang Fu (Cyperus rotundus L.) (24.9%)

Chaihu Shugan decoction

Ganmai Dazao decoction Guipi decoction

Wendan decoction/ Shiwei Wendantang decoction Banxia Houpu decoction

Chaihujialonggumuli decoction

Tiaoqi hypoglycemic pills

Yipi hypoglycemic pills

Tangshenkang

Frequently used single Chinese herb

345 CHM formulas (53.0%); the second most frequently used single herb was Bai Shao (Paeonia lactiflora Pall.) (42.3%), followed by Fu Ling (Poria Cocos (Schw) Wolf.) (41.7%). The number of single herbs in the formulas ranged from 1 to 59, with a mean of 13. Twohundred and ninety-four (85.2%) of the 345 formulas were prepared in decoction, twenty-two (6.4%) in capsule, and twenty-nine (8.4%) in pills or pellets. The CHM was taken 1e3 times per day, with a median of 2 times per day. The treatment ranged from 10 to 180 days, with an average of 49.1 days. 3.3. Quality assessment 3.3.1. Assessment by the modified Jadad Scale Table 1 summarizes the study quality of the 296 included studies. Two-hundred seventy-five studies (92.9%) were rated as low quality RCTs (modified Jadad scores  2) and the remaining 21 were moderate in methodological quality. Table 3 presents the characteristics of the 21 studies with better methodological quality, of which we had examined in details. All the 21 studies were reported as randomized trials, but in three studies (Studies 3, 5 and

N (%)

18), the appropriate randomization method was not described. Studies 3 and 5 only used alternation as randomization, while Study 18 did not report the randomization method. Nine studies fulfilled the criteria of double-blind study design and described an appropriate method (Studies 1e3, 5, 6, 8, 11 and 17e18). Although Studies 4, 15 and 19 described that double-blind study design was used, they did not report the method. All studies, except Study 17, had adequate description of dropouts. 3.3.2. Assessment by Cochrane's risk of bias assessment Eleven of the 21 studies (Studies 3, 5, 7, 9e10, 12e14, 16 and 20e21) had at least one domain rated as high risk of bias despite having a modified Jadad score
3. Only five studies had low risk of bias in all domains (Studies 1e2, 6, 8 and 11). The random sequence generation in Study 18 was rated as unclear as there was no report of the randomization method. Study 3 and 5 used alternation as randomization, so the study was rated to have a high risk of bias in random sequence generation and allocation concealment. Of the 18 studies which reported an adequate randomization method, only six studies (Studies 1e2, 6e8 and 11) used an adequate

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175 Table 3 Cochrane risk of bias assessment and modified Jadad score of the 21 moderatequality RCTs on Chinese herbal medicine for depression. No.

1st Author (year)

Cochrane risk of biasa

Modified Jadad scores (Total score)b

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Akhondzadeh et al. (2005) Moshiri et al. (2006) Sun et al. (2009) Yang et al. (2002) Zhang et al. (1998) Akhondzadeh et al. (2004) Bao et al. (2012) Basti et al. (2007) Chen and Zhang (2010) Guo et al. (2012) Noorbalaa et al. (2005) Shen et al. (2004) Ta and Wang (2008) Tao (2006) Yang et al. (2012) Lin et al. (2008) Li et al. (2007) Luo et al. (1999) Yi et al. (2010) Fu et al. (2008) Sun et al. (2012)

L, L, L, L, L, L, L L, L, L, L, L, L, L H, H, L, L, L, L, U L, U, L, U, L, L, L H, H, L, L, L, L, U L, L, L, L, L, L, L L, L, H, H, L, L, L L, L, L, L, L, L, L L, U, H, H, L, L, L L, U, H, H, L, L, L L, L, L, L, L, L, L L, U, H, H, L, L, L L, U, H, H, L, L, L L,U, H, H, L, L, U L, U, U, U, L, L, L L, U, H, H, L, H, L L, U, L, L, L, L, L U, U, L, L, L, L, L L, U, L, U, L, L, L L, U, H, H, L, L, L L, U, H, H, L, L, L

1,1,1,1,1 1,1,1,1,1 1,0,1,1,1 1,1,1,0,1 1,0,1,1,1 1,1,1,1,1 1,1,0,0,1 1,1,1,1,1 1,1,0,0,1 1,1,0,0,1 1,1,1,1,1 1,1,0,0,1 1,1,0,0,1 1,1,0,0,1 1,1,0,0,1 1,1,0,0,1 1,1,1,1,0 1,0,1,1,1 1,1,1,0,1 1,1,0,0,1 1,1,0,0,1

(5) (5) (4) (4) (4) (5) (3) (5) (3) (3) (5) (3) (3) (3) (3) (3) (4) (4) (4) (3) (3)

a

High risk (H), unclear risk (U) or low risk (L) of bias in random sequence generation, allocation concealment, blinding of participants, blinding of outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias. b Adequate (1) or inadequate (0) whether the study is described as randomized, the study has an appropriate randomization method, the subjects are blinded to intervention, the evaluators are blinded to intervention, and the study has description of withdrawals and dropouts.

concealment method, while the other studies were rated to have an unclear risk of bias in allocation concealment. In nine studies (Studies 7, 9e10, 12e14, 16 and 20e21), neither the participants nor the evaluators were blinded, hence there was a high risk of bias in blinding. Although Studies 4, 15 and 19 reported using a doubleblind study design, none of them were truly double-blinded; Studies 4 and 19 were rated to have a low risk in subjectblinding, but there was an unclear risk in evaluator-blinding because the assessors might not be blind to treatment allocation, whereas Study 15 did not have adequate description of how the participants and assessors were blinded; hence, there was an unclear risk of bias in blinding. All 21 studies, except Studies 3, 5 and 14, had no information whether baseline differences in subject characteristics were present or not; hence they were rated to have an unclear risk in other sources of bias.

169

Depression Scale, Zung Self-Rating Anxiety Scale, and a depression inventory developed by the Hunan Medical University (n ¼ 1). The HDRS17/24 was the most commonly used outcome measure. Adverse events were assessed in 13 of the 21 studies; three by the Treatment Emergent Symptom Scale (Studies 19e21), two by the Asberg Side Effect Rating Scale (Studies 17 and 18), and the remaining eight studies were based on self-report. Eighteen studies used standardized formula and 3 used individualized treatment. Saffron extract was used in five studies, followed by Xiaoyao decoction (Free and Easy Wanderer Plus) or its modification, which were used in four studies; and Jieyu pill was used by two studies. The other studies used different formulas. In 15 of the 21 included studies, CHM was given for six weeks, and it was eight weeks in four studies and 60 days in two studies. 3.5. Efficacy The 21 studies with modified Jadad score
3 were included in meta-analysis. 3.5.1. CHM vs. placebo (Studies 1e5) Two studies conducted in Iran found that depressed patients treated with saffron (Crocus sativus L.) had significantly greater reduction in HDRS score than those on placebo (Studies 1 and 2) (MD: 7.97, 95% CI: 10.25 to 5.70, P < 0.00001, I2 ¼ 0%). Studies 3, 4 and 5 examined the antidepressant effects of Shugan Jieyu capsule, Shujiele wutang pill, and Xiao Yao pill, respectively. Pooled analysis found that CHM had significantly higher effective rate than placebo (RR: 2.99, 95% CI: 2.18 to 4.10, P < 0.00001, I2 ¼ 0%) but there was no significant difference in end-point HDRS score. Details of the pooled analyses are presented in Tables 5 and 6. 3.5.2. CHM vs. antidepressants (Studies 6e15 and 20e21) Of the 12 studies comparing CHM with antidepressants, seven studies used fluoxetine 10e40 mg/day (Studies 7e9, 11, 13, 15 and 21), while the other studies used other selective serotonin reuptake inhibitors, selective serotonin and norepinephrine reuptake inhibitors, tricyclic or tetracyclic antidepressants, or deanxit, a combination of flupentixol and melitracen as antidepressants. Pooled analysis showed that there was no significant difference between CHM and antidepressants in terms of effective rate (RR: 1.00, 95% CI: 0.94 to 1.07, P ¼ 1.00, I2 ¼ 42%), end-point HDRS score (SMD: 0.01, 95% CI: 0.28 to 0.30, P ¼ 0.95, I2 ¼ 80%) and HDRS change score (MD: 2.05, 95% CI: 0.09 to 4.18, P ¼ 0.06, I2 ¼ 0%) (Tables 7 and 8).

3.4. Description of the studies with modified Jadad score
3 Table 4 summaries the characteristics of the 21 studies with better methodological quality. A total of 2871 participants were recruited in the studies; 52.7% was female, the mean age was 39.9 years and the range from 16 to 78 years. Nineteen of the 21 studies were 2-arm and the remaining two were 3-arm studies. Seventeen studies recruited subjects with depression in general; two studies were on post-stroke depression, one study on post-partum depression, and one study on neurotic depression. The included studies used different diagnostic systems and rating scales for subject selection. The most commonly used rating scale was the 17item or 24-item Hamilton Depression Rating Scale (HDRS17/24), which was used in 17 studies; a total score from eight to 28 was set as inclusion criteria. Other diagnostic criteria used in the studies include the CCMD-2-R/3 (n ¼ 11), DSM-IV (n ¼ 8), TCM textbooks (n ¼ 4), Zung Self-Rating Depression scale (n ¼ 3), International Classification of Diseases, Tenth Edition (n ¼ 2), Edinburgh Postnatal

3.5.3. Adjunctive CHM with antidepressants vs. antidepressants alone (Studies 16 and 20) In Studies 16 and 20, paroxetine 20 mg/day and deanxit 2 tablets/day were used, respectively, as antidepressants. Pooled analysis showed that adjunctive CHM with antidepressants was significantly better than antidepressants alone in effective rate (RR: 1.17, 95% CI: 1.04 to 1.32, P ¼ 0.01, I2 ¼ 0%), but there was no difference in end-point HDRS score (SMD: 1.31, 95% CI: 3.24 to 0.63, P ¼ 0.19, I2 ¼ 96%). 3.5.4. Adjunctive CHM with placebo antidepressants vs. adjunctive placebo CHM with antidepressants(Study 17) Maprotiline 25e250 mg/day was the antidepressant used in the study. Statistical analysis showed that there was no significant group difference in end-point HDRS score (MD: 2.45, 95% CI: 1.62 to 6.52, P ¼ 0.24).

No. 1st Author (year) CHM vs. Placebo 1 Akhondzadeh et al. (2005) 2 Moshiri et al. (2006) 3 Sun et al. (2009) 4 5

Yang et al. (2002) Zhang et al. (1998)

Country/type of case

Mean age Diagnostic system/ (range)/% female inclusion criteria (cutoff)

Design

Sample size (CHM/control)

Treatment/co-intervention (duration)

Control intervention

Primary and other outcome measures

Results reported

Iran/outpatients with depression Iran/outpatients with depression China/in and outpatients with depression China/patients with depression China/outpatients with neurotic depression

36.3 (NR)/55.0%

DSM-IV, HDRS17 (
18)

40 (20/20)

Saffron (6 weeks)

Placebo

DSM-IV, HDRS17 (>18)

40 (20/20)

Saffron (6 weeks)

Placebo

38.4 (NR)/86.9%

CCMD3, HDRS17 (17e28)

120 (80/40)

Shugan Jieyu capsule (6 weeks)

Placebo

HDRS17 change score HDRS17 change score End-point HDRS17

CHM > placebo

35.7 (NR)/57.5%

2 parallel arms (CHM; placebo) 2 parallel arms (CHM; placebo) 2 parallel arms (CHM; placebo)

36.2 (NR)/63.2%

DSM-IV, HDRS17 (
8), TCM textbook DI

68 (36/32)

Placebo

110 (58/52)

Shujiele wutang pill (6 weeks) Xiaoyao pill (60 days)

Placebo

30 (15/15)

Saffron (6 weeks)

Imipramine 100 mg/d

HDRS17

CHM ¼ AD

100 (50/50)

Zhuyang shuxin decoction (6 weeks) Saffron (8 weeks)

Fluoxetine 20 mg/d

Effective rate

CHM > AD

Fluoxetine 20 mg/d

HDRS17 change, effective rate based on HDRS Effective rate based on HDRS and TCM symptoms, NFDS

CHM ¼ AD

40.1 (20e56)/46.4%

CHM > placebo CHM > placebo

2 parallel arms (CHM; AD) CCMD3, TCM textbook 2 parallel arms (CHM; AD) DSM-IV, HDRS17 (19e25) 2 parallel arms (CHM; AD) CCMD3, HDRS (8e23), TCM textbook

2 parallel arms (CHM þ RT; AD þ RT)

60 (30/30)

Yangqi Huoxue decoction (6 weeks)

Fluoxetine 10e20 mg/d

CCMD3, HDRS17 (>17)

2 parallel arms (CHM; AD)

160 (80/80)

Fluoxetine/Paroxetine/ Citalopram 20 mg/d, Sertraline 50 mg/d or Venlafaxine 75 mg/d

Effective rate based on HDRS17 and TCM symptoms

CHM ¼ AD

DSM-IV, HDRS17 (>18)

2 parallel arms (CHM; AD) 2 parallel arms (CHM; AD)

40 (20/20)

Xuefu Zhuyu decoction, Tianwang Buxin pill, Zishui Qinggan decoction, Qifu decoction, or Liuwei Dihuang pill (8 weeks) Saffron (6 weeks)

Fluoxetine 20 mg/d

HDRS17 change

CHM ¼ AD

60 (30/30)

Jieyu pill (6 weeks)

Maprotiline 100e250 mg

CHM ¼ AD

Chen and Zhang (2010)

10

Guo et al. (2012)

56.2 (NR)/45.0% China/inpatients and outpatients with post-stroke depression China/outpatients 34.8 (NR)/53.8% with depression

11

Noorbalaa et al. (2005) Shen et al. (2004)

Iran/outpatients with depression China/patients with depression

36.9 (18e55)/50.0% 33.0 (16e57)/58.3%

DSM-IV, HDRS17 (
18)

Effective rate based on HDRS17, CGI Effective rate based on DI

CHM > placebo

Iran/outpatients 34.0 (NR)/43.3% with depression China/outpatients 40.4 (NR)/68.0% with depression Iran/outpatients 34.8 (NR)/47.5% with depression

9

12

2 parallel arms (CHM; placebo) 2 parallel arms (CHM; placebo)

CHM > placebo

CHM > AD

13

Ta and Wang (2008)

2 parallel arms (CHM þ RT; AD þ RT)

48 (24/24)

Chaihu jialonggu muli decoction (60 days)

Fluoxetine 20 mg/d

14

Tao (2006)

2 parallel arms (CHM; AD)

86 (41/45)

Jieyu pill (6 weeks)

Venlafaxine 75e225 mg/d

Effective rate based on HDRS24, CGI, SAS, SDS Effective rate based on HDRS17 and TCM syndrome diagnostic system Effective rate based on HDRS17

15

Yang et al. (2012)

2 parallel arms (CHM; AD)

1024 (512/512) Qican Fukang capsule (6 weeks)

Fluoxetine 20e40 mg/d

Effective rate based on HDRS17, CGI

CHM > AD

2 parallel arms (CHM þ AD; AD)

68 (37/31)

Free and Easy Wanderer Plus (6 weeks)

Paroxetine 20 mg/d

Effective rate based on HDRS24

CHM þ AD > AD

Danzhi Xiaoyao decoction (6 weeks)

Maprotiline 25e250 mg/d and placebo

ASBS, HDRS24, SAS, SDS

CHM þ placebo > AD þ placebo

Adjunctive CHM 16 Lin et al. (2008) Adjunctive CHM 17 Li et al. (2007)

CCMD3, HDRS24 (
20), ICD-10, SDS,

40 (20/20)

65.0 CCMD3, HDRS (NR), China/in and (NR)/45.8% TCM textbook outpatients with post-stroke depression 32.43 HDRS17 (
18), ICD-10 China/in and (NR)/57.0% outpatients with depression 38.8 (NR)/39.9% DSM-IV China/in and outpatients with depression with AD vs. AD DSM-IV, EPDS, HDRS24 China/outpatients 26.9 (NR)/100% with post-partum (
20), depression with placebo vs. Adjunctive placebo CHM with AD China/patients 39.2 (NR)/51.5% CCMD3, HDRS24 (
20), with depression ICD-10, SAS, SDS

Adjunctive CHM with AD vs. Adjunctive placebo CHM with AD 18 CCMD2-R, HDRS24 (
20)

66 (33/33) 2 parallel arms (CHM þ placebo; AD þ placebo) 228 (114/114)

Shuxue ning (8 weeks)

CHM þ RT ¼ AD þ RT

CHM ¼ AD

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

CHM vs. AD 6 Akhondzadeh et al. (2004) 7 Bao et al. (2012) 8 Basti et al. (2007)

170

Table 4 RCTs of Chinese herbal medicine for depression with modified Jadad score
3.

CCMD3, TCM textbook 47.5 (16e74)/51.3% Sun et al. (2012) 21

AD, antidepressant; ASBS, Asberg side effect rating scale; CCMD, Chinese Classification of Mental Disorder; CGI, Clinical Global Impression; CHM, Chinese herbal medicine; DI: Depression Inventory from Hunan Medical University; HDRS, Hamilton Depression Rating Scale; NFDS, Neural Function Deficient Scale; NR, not reported; RT, routine treatment; SAS, Zung Self-rating Anxiety Scale; SDS, Zung Self-rating Depression Scale; TCM, traditional Chinese medicine; TESS, Treatment Emergent Symptoms Scale.

CHMa ¼ AD > CHMb Effective rate based on HDRS, SDS, TCM symptoms, TESS Fluoxetine 20 mg/d CHMa: Shuyu decoction, CHMb: modified Free and Easy Wanderer Plus (6 weeks) 161 (53/52/56)

CHM þ AD > CHM ¼ AD Effective rate based on HDRS, TESS Wuling capsule (6 weeks) 3-arm design 20 Fu et al. (2008)

China/in and outpatients with depression China/outpatients with depression

52.5 (45e78)/54.2%

CCMD3, HDRS (
9)

3 parallel arms (CHM þ AD; CHM; AD) 3 parallel arms (CHMa; CHMb; AD)

120 (40/40/40)

Deanxit 2 tablets/d

CHM þ AD > AD þ placebo Chaihu Xiaoyao decoction (8 weeks) 43.6 (NR)/54.7% China/in and outpatients with depression Yi et al. (2010) 19

Luo et al. (1999)

China/patients with depression

37.5 (18e62)/43.9%

CCMD3, HDRS17 (
17)

2 parallel arms (CHM þ AD; AD þ placebo) 2 parallel arms (CHM þ AD; AD þ placebo)

190 (92/98)

Amitriptyline 20 mg/d and placebo

Effective rate based on HDRS24, ASBS, CGI Paroxetine 20e40 mg/d Effective rate based and placebo on HDRS17, CGI, TESS

CHM þ AD > WM þ AD

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

171

3.5.5. Adjunctive CHM with antidepressants vs. adjunctive placebo CHM with antidepressants (Studies 18e19) Study 18 found that adjunctive CHM with amitriptyline 20 mg/ day had significantly higher effective rate and greater reduction in HDRS score than placeboeamitriptyline combination. Similarly, Study 19 found that adjunctive CHM with paroxetine 20e40 mg/ day was more efficacious than placeboeparoxetine combination. Pooled analysis showed that adjunctive CHM with antidepressants had significantly higher effective rate (RR: 1.18, 95% CI: 1.06 to 1.30, P ¼ 0.002, I2 ¼ 0%), but it was not possible to perform pooled analysis of HDRS score due to incompatible data in the two studies. 3.6. Adverse events Adverse events were reported in 92 (31.2%) of the 296 reviewed studies, and in 72 (78.2%) of the 92 studies, the frequency of adverse events was also provided. Table 7 summarizes the most frequently reported adverse events. Meta-analysis showed that subjects taking CHM were less likely to report adverse events than those taking antidepressants (n ¼ 29, pooled RaR: 0.23, 95% CI: 0.16 to 0.33, P < 0.00001, I2 ¼ 59%). Almost all the commonly reported adverse events, except diarrhea, were less frequent in participants taking CHM (Table 7). Another meta-analysis showed that CHMantidepressant combination was associated with less adverse events than antidepressant alone (n ¼ 38, pooled RaR: 0.43, 95% CI: 0.35 to 0.52, P < 0.00001, I2 ¼ 64%) and there was no significant difference between CHM and placebo in adverse events (n ¼ 3, pooled RaR: 1.29, 95% CI: 0.86 to 1.95, P ¼ 0.22, I2 ¼ 61%). Anxiety, reduced appetite and dry mouth were the adverse events occurring more than 3% in people taking CHM in the studies comparing CHM with antidepressants (Table 9). 4. Discussion This is the first systematic review aiming to derive an overall picture of the efficacy and safety of CHM for depression in better quality RCTs and to examine the ingredients of the herbal formulas used for the treatment of depression. The results from the 296 included RTCs and meta-analyses of the studies with better methodological quality seemingly suggested that CHM was superior to placebo and as effective as antidepressants for treating depression. However, most of the RCTs we had reviewed scored 2 or less by the modified Jadad scale and were deemed to be of low quality. Owing to the small number of studies using the same herbal formula and the heterogeneity in subject characteristics, dosage and type of antidepressants and outcome assessment, the results should be treated with caution. Hence, it is premature to conclude that CHM is indeed an alternative treatment for depression. The review found that a majority of the RTCs had poor methodological quality. The procedures of randomization and allocation concealment were not properly described in many studies. Blinding of the patients and outcome assessors was rarely used. Placebo design is a challenge to RCTs of CHM. Although decoction is the favored method of extraction of herbal material, herbal formulas and placebo can be prepared in pills, capsules, or granules to achieve blinding. Further studies should follow the good manufacturing practice guidelines (World Health Organization, 2006) and the standards in the Chinese pharmacopoeia (Chinese Pharmacopoeia Commission, 2010) to ensure the pills, capsules, and granules contain the same amount of active ingredients as in the CHM decoction. It is noted that most of the reviewed studies did not include the Latin scientific names of individual Chinese herbal medicine. The lack of precise use of botanical scientific nomenclature may lead to taxonomic ambiguity (Rivera et al., 2013), making

172

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

Table 5 Chinese herbal medicine (CHM) vs. placebo as treatment for depression. Study

CHM

Placebo

N improveda Effective rate Sun et al., 2009

78

Weight

N improveda

Risk ratio, Fixed, 95% CI

More effective with placebo

Total N

8

34

31.8%

2.56 [1.36, 4.82]

Yang et al., 2002 32 35 5 Zhang et al., 1998 53 58 17 Total (95% CI) 171 Total events 132 30 2 Heterogeneity: Chi ¼ 1.26, df ¼ 2 (P ¼ 0.53); I2 ¼ 0% Test for overall effect: Z ¼ 6.80 (P < 0.00001)

24 52 110

16.9% 51.2% 100%

4.39 [2.00, 9.64] 2.80 [1.88, 4.16] 2.99 [2.18, 4.10]

a

47

Total N

More effective with CHM

Improved, defined as the proportion of subjects who had at least 30% reduction in depression score.

Table 6 Chinese herbal medicine (CHM) vs. placebo as treatment for depression. Study

CHM Mean

Placebo

Weight

Mean difference, random, 95% CI

SD

Mean

SD

4.16

13.29

4.94

53.9%

3.78 [5.68, 1.88]

Yang et al., 2002 6.30 7.70 17.90 9.80 Total (95% CI) Heterogeneity: Tau2 ¼ 27.26; Chi2 ¼ 9.22, df ¼ 1 (P ¼ 0.002); I2 ¼ 89% Test for overall effect: Z ¼ 1.89 (P ¼ 0.06)

46.1%

11.60 [16.28, 6.92] 7.39 [15.03, 0.25]

End-point HDRS scorea Sun et al., 2009

9.51

More effective with CHM

More effective with placebo

Mean Difference, Fixed, 95% CI

HDRS change score 12.20

Akhondzadeh et al., 2005

4.67

Moshiri et al., 2006 14.01 5.53 Total (95% CI) Heterogenity: Chi2 ¼ 0.64, df ¼ 1 (P ¼ 0.42); I2 ¼ 0% Test for overall effect: Z ¼ 6.87 (P < 0.00001)

5.10

4.71

53.1%

7.10 [10.22, 3.98]

5.05

4.63

46.9% 100%

8.96 [12.28, 5.64] 7.97 [10.25, 5.70]

HDRS, Hamilton Depression Rating Scale. a No baseline difference in HDRS based on authors' calculation.

it difficult to assess treatment integrity and replicate the CHM treatment. In addition, most of the trials only focused on effective rate and short-term outcome and their sample size was usually small. The large number of low-quality studies identified in this review suggested that the researchers might not be aware of the crucial

components of RCTs and the publishing journals had not placed sufficient emphasis on the standards of reporting trials. A series of papers has been published on improving the quality of RCTs in CHM (Bian et al., 2006a, 2006b; Leung et al., 2006; Sarris, 2012). The key recommendations include: 1) investigators conducting RCTs should have formal training about clinical trial design; 2) registration of clinical

Table 7 Chinese herbal medicine (CHM) vs. antidepressants (AD) as treatment for depression. Study

CHM N improveda

Effective Rate Bao et al., 2012

46

AD Total N 50

N improveda 42

Basti et al., 2007 15 19 17 Chen and Zhang, 2010 25 29 17 Fu et al., 2008 34 39 32 Guo et al., 2012 43 66 42 Shen et al., 2004 27 28 29 Sun et al., 2012 95 105 48 Ta and Wang, 2008 20 24 19 Tao 2006 21 41 37 Yang et al., 2012 389 445 389 Total (95% CI) 846 Total events 715 672 2 2 Heterogeneity: Tau ¼ 0.00; Chi ¼ 15.63, df ¼ 9 (P ¼ 0.08); I2 ¼ 42% Test for overall effect: Z ¼ 0.01 (P ¼ 1.00)

a

Weight Total N 50

11.8%

19 27 37 64 29 56 24 45 438 789

4.6% 3.5% 9.3% 5.5% 17.7% 14.2% 4.7% 3.4% 25.3% 100%

Risk ratio, random, 95% CI

More effective with AD

1.10 [0.95, 1.27] 0.88 1.37 1.01 0.99 0.96 1.06 1.05 0.62 0.98 1.00

[0.67, [0.99, [0.85, [0.77, [0.88, [0.93, [0.80, [0.45, [0.94, [0.94,

1.17] 1.89] 1.20] 1.27] 1.06] 1.19] 1.38] 0.86] 1.03] 1.07]

Imporved, often defined as the proportion of subjects who had at least 30% reduction in the depression rating scale score.

More effective with CHM

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

173

Table 8 Chinese herbal medicine (CHM) vs. antidepressants (AD) as treatment for depression. Study

CHM Mean

AD SD

Mean

Weight SD

Standardized mean difference, random, 95% CI

More effective with CHM

More effective with AD

End-point HDRS scorea Chen and Zhang, 2010 9.76 6.58 Fu et al., 2008 10.54 3.62 Guo et al., 2012 10.42 2.36 Shen et al., 2004 9.71 4.78 Sun et al., 2012 13.46 4.75 Tao 2006 13.54 33.21 Yang et al., 2012 19.40 4.20 Total (95% CI) 2 2 Heterogeneity: Tau ¼ 0.11, Chi ¼ 29.78, df ¼ 6 Test for overall effect: Z ¼ 0.06 (P ¼ 0.95)

14.55 13.70 9.77 8.90 12.10 9.23 18.00

10.03 3.63 2.89 6.68 2.17 4.81 5.40

11.8% 12.9% 15.3% 12.0% 15.6% 13.8% 18.7% 100% (P < 0.0001); I2 ¼ 80%

Basti et al., 2007 12.00 4.10 13.50 Noorbalaa et al., 2005 12.20 4.67 15.00 Total (95% CI) Heterogeneity: Chi2 ¼ 0.35, df ¼ 1 (P ¼ 0.56); I2 ¼ 0% Test for overall effect: Z ¼ 1.88 (P ¼ 0.06)

4.91 5.88

HDRS change score

0.56 [1.10, 0.03] 0.86 [1.33, 0.39] 0.25 [0.10, 0.59] 0.14 [0.38, 0.66] 0.33 [0.01, 0.66] 0.18 [0.24, 0.61] 0.29 [0.16, 0.42] 0.01 [0.28, 0.30]

Mean Difference, Fixed, 95% CI

HDRS, Hamilton Depression Rating Scale.

a

57.9% 42.1% 100%

1.50 [1.30, 4.30] 2.80 [0.49, 6.09] 2.05 [0.09, 4.18]

No baseline difference in HDRS based on authors' calculation.

trials and publishing their protocols in recognized trial registration platforms, such as the Chinese clinical trial registration; 3) collaboration between researchers in different fields; 4) understanding the rationale for selecting different types of control groups; 5)

development and implementation of good agricultural practice, good manufacturing practice, and good clinical practice in CHM research; and 6) a revised CONSORT checklist should be used in reporting RCTs of CHM. Taken together, our findings suggest that further studies with

Table 9 Risk ratio of the 10 most frequently reported adverse events in studies comparing Chinese herbal medicine (CHM) with antidepressants (AD), in studies comparing CHM with placebo, and in studies comparing adjunctive CHM with AD with AD alone. Adverse events

Total events/group total CHM

Anxiety 14/386 Reduced appetite 13/482 Dry mouth 12/397 Constipation 8/342 Headache 8/300 Nausea 8/413 Increased appetite 7/118 Diarrhea 5/241 Insomnia 4/458 Abdominal bloating 3/136 Incidence of any adverse event Adjunctive CHM with AD Dry mouth 90/614 Dyspepsia 39/286 Nausea 37/538 Dizziness 37/538 Constipation 35/587 Headache 33/415 Tachycardia 31/505 Blurred vision 20/431 Diarrhea 17/296 Weary limbs 15/231 Incidence of any adverse event CHM Nausea 12/116 Reduced appetite 8/116 Anxiety 7/38 Headache 6/38 Dizziness 5/78 Tachycardia 5/97 Increased appetite 5/19 Stomach ache 4/19 Tremor 3/19 Hypomania 2/38 Incidence of any adverse event

Risk ratio (95% CI)

No. of studies

I2 (%)

0.40 (0.24, 0.67) 0.35 (0.20, 0.59) 0.19 (0.11, 0.32) 0.18 (0.10, 0.34) 0.25 (0.13, 0.47) 0.24 (0.13, 0.45) 0.36 (0.06, 2.09) 1.39 (0.43, 4.46) 0.21 (0.11, 0.41) 0.17 (0.05, 0.62) Pooled rate ratio (95% CI) 0.23 (0.16, 0.33)

11 12 10 8 9 11 4 4 12 2 29

0% 21% 37% 43% 0% 0% 65% 15% 1% 55% 59%

0.57 (0.46, 0.71) 0.43 (0.31, 0.61) 0.41 (0.32, 0.65) 0.45 (0.31, 0.64) 0.43 (0.31, 0.61) 0.60 (0.42, 0.87) 0.58 (0.39, 0.87) 0.43 (0.27, 0.70) 0.72 (0.41, 1.27) 0.61 (0.18, 2.05) Pooled rate ratio (95% CI) 0.43 (0.35, 0.52)

21 8 16 15 16 11 13 12 7 5 38

8% 43% 44% 3% 22% 36% 0% 44% 45% 58% 64%

2.55 (0.80, 8.18) 1.20 (0.42, 3.43) 2.04 (0.57, 7.25) 1.73 (0.47, 6.35) 1.09 (0.22, 5.34) 1.68 (0.41, 6.83) 4.21 (0.55, 32.43) 1.79 (0.37, 8.57) 2.68 (0.31, 23.43) 0.86 (0.16, 4.77) Pooled rate ratio (95% CI) 1.29 (0.86, 1.95)

3 3 2 2 1 2 1 1 1 2 3

0% 0% 0% 0% NA 0% NA NA NA 0% 61%

AD 38/353 43/444 77/399 60/349 38/292 40/340 17/94 3/227 42/453 14/94 AD 148/666 71/251 93/553 80/503 82/562 54/397 53/487 45/418 21/275 32/224 Placebo 3/67 5/67 3/33 3/33 2/34 2/51 1/16 2/17 1/17 2/33

174

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175

improved methodology are warranted in order to accurately determine the efficacy and safety of CHM for depression. We found that nearly 100 different Chinese herbal formulas were developed and used to treat depression. The great variety of possible herbal combination is due in part to the complexity of TCM diagnosis and the practitioners' personal experiences. Based on patients' symptoms and signs, TCM practitioners analyze the patterns of bodily disharmony and prescribe herbal formulas according to the TCM diagnosis. Xiao Yao decoction was the most frequently examined formula, followed by Chaihu Shugan decoction and Ganmai Dazao decoction. Both Xiao Yao decoction and Chaihu Shugan decoction are for liver depression and they both contain Chai Hu (B. chinense DC.), Dang Gui (Angelica sinensis (Oliv.) Diels), Bai Shao (P. lactiflora Pall.) and Gan Cao (G. uralensis Fisch.). Xiao Yao decoction can be used specifically for those accompanied with spleen and blood deficiency in which dizziness, tiredness and loss of appetite may be present; whereas Chaihu Shugan decoction is for those accompanied with qi stagnation, in which belching and gastric regurgitation may be found. Ganmai Dazao decoction contains only three herbs, Gan Cao (G. uralensis Fisch.), Xiao Mai (Triticum aestivum L.) and Da Zao (Ziziphus jujuba Mill.) and it is used to treat “Zangzao” (visceral irritation), in which female patients presented with low mood, a tendency to cry, possession-like behavior and frequent yawning due to heart deficiency (Chen and Zhang, 2000). Due to the small number of high-quality RCTs and a wide variety of herbal combination, it is premature to make any conclusion on the merits of the different formulas and single herbs. We found that standardized CHM formulas and individualized CHM treatment based on TCM pattern diagnosis were both commonly used in the included studies. However, no head-to-head comparison between individualized and standardized CHM for depression has been performed. Further studies are warranted to examine the usefulness of individualized CHM treatment based on TCM pattern diagnosis. However, it is also important to overcome the limitations of the TCM diagnostic process, including the lack of standardization in terminology and disagreement of pattern diagnosis among Chinese medicine practitioners. Pharmacologic actions of CHM for depression have been explored both in vitro and in vivo studies. Studies showed that Xiao Yao decoction and Chaihu Shugan decoction could alleviate the dysregulation of corticotrophin-releasing hormone and adrenocorticotropic hormone induced by chronic mild stress (Hu et al., 2010; Wu et al., 2006). Another study showed that Chuan Xiong (Ligusticum striatum DC.), one of the 10 most frequently used Chinese herbs for depression, and Huang Qin (Scutellaria baicalensis Georgi), an ingredient of Chaihujialonggumuli decoction and Tiaoqi hypoglycemic pills, could reverse depression-like behavior, weight loss and corticosterone levels and induce hippocampal neurogenesis (Pao et al., 2012). Two other commonly used ingredients in the CHM for depression, Dang Gui (A. sinensis (Oliv.) Diels) and Chai Hu (B. chinense DC.), were shown to act on receptor binding sites of several central neurotransmitters including GABAA and dopamine receptors (Chen et al., 2008; Liao et al., 1995). Additionally, Xiao Yao decoction has been found to have anxiolytic and sedative effects (Mizowaki et al., 2001; Yi et al., 2007). Ganmai Dazao decoction, the third commonly used formula, has been shown to reverse depression-like behavior, via pathways involving glutamate, Nmethyl-D-aspartate receptor, pro-inflammatory cytokines, hypothalamic-pituitary-adrenal axis, nitric oxide and brain-derived neurotrophic factor (Hu, 2011; Tong et al., 2005; Zhang et al., 2012). The overall results seem to support the pharmacologic actions of CHM for depression; however, there are still limited data, particularly in humans, therefore, further studies are definitely needed. We found that the frequency of adverse events was similar between CHM and placebo; however, only three studies were

available. Adverse events were more common in people taking antidepressants compared to those taking CHM, but the addition of CHM to antidepressants seemed to reduce the frequency their side effects. The common adverse events associated with CHM were anxiety, reduced appetite and dry mouth. No serious adverse events were reported in the reviewed studies. However, a majority of the studies had not used standardized rating scales for treatment emergent adverse events. Furthermore, the risks of adverse interactions between Chinese herbs and central nervous system depressants are not known. Previous studies have found that Chinese herbal agents can alter the pharmacokinetics of many drugs, which may cause serious side effects (Fugh-Berman, 2002; Izzo and Ernst, 2009). Future studies on the efficacy and safety of combinations of Chinese herbal agent and Western medication are required. The long-term safety of Chinese herbs, the effects of overdose and their use in pregnancy and during lactation were not clearly recorded and this merits further investigation. 5. Conclusion Although the overall results from the 296 included RCTs and the meta-analyses of the 21 studies with modified Jadad scale score of 3 or above showed that CHM was more effective than placebo and as effective as antidepressants in the treatment of depression, the many shortcomings of the reviewed RCTs and the lack of adequate safety data disallow a definite conclusion on the efficacy and safety of CHM for depression. Future studies with improved methodology targeted at some potentially efficacious and safe herbal formulas will enrich our knowledge about the use, value and risks of CHM for depression. Role of funding source The present study was supported by the Chinese Medicine Research Fund (Project no.: HACMR001) from Hospital Authority of Hong Kong. Contributors WF Yeung and KF Chung conceived the primary research question for the study. WF Yeung, KF Chung, ET Ziea and BF Ng were involved in the study conception and design. KY Ng and YM Yu were responsible for data extraction. WF Yeung, KF Chung and KY Ng interpreted the results and drafted the initial manuscript. WF Yeung and KF Chung provided input on interpretation of results and provided critical revision to the manuscript. All authors read and approved the final manuscript. Conflict of interest No competing financial interests exist. Acknowledgment We would like to thank the Hospital Authority of Hong Kong for the fund (HACMR001). References Akhondzadeh S, Fallah-Pour H, Afkham K, Jamshidi AH, Khalighi-Cigaroudi F. Comparison of Crocus sativus L. and imipramine in the treatment of mild to moderate depression: a pilot double-blind randomized trial [ISRCTN45683816]. BMC Complement Altern Med 2004;4:12. Akhondzadeh S, Tahmacebi-Pour N, Noorbala AA, Amini H, Fallah-Pour H, Jamshidi AH, et al. Crocus sativus L. in the treatment of mild to moderate

W.-F. Yeung et al. / Journal of Psychiatric Research 57 (2014) 165e175 depression: a double-blind, randomized and placebo-controlled trial. Phytother Res 2005;19:148e51. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 4th ed. Washington DC: American Psychiatric Association; 1994. Bao ZX, Tian Q, Zhou W, Sun W, He GP. Clinical observation of influence of Zhuyang Shuxin recipe on TCM syndrome of patients with depression with syndrome of liver and kidney yang deficiency. Chin Arch Tradit Chin Med 2012;30:116e8 [In Chinese]. Basti A, Moshiri E, Noorbala AA, Jamshidi AH, Abbasi SH, Akhondzadeh S. Comparison of petal of Crocus sativus L. and fluoxetine in the treatment of depressed outpatients: a pilot double-blind randomized trial. Prog NeuroPsychopharmocol Biol Psychiatry 2007;31:439e42. Baxter AJ, Scott KM, Ferrari AJ, Norman RE, Vos T, Whiteford A. Challenging the myth of an “epidemic” of common mental disorders: trends in the global prevalence of anxiety and depression between 1990 and 2010. Depress Anxiety 2014;31:506e16. Bian ZX, Moher D, Dagenais S, Li YP, Liu L, Wu TX, et al. Improving the quality of randomized controlled trials in Chinese herbal medicine, part II: control group design. Chin J Integr Med 2006a;4:130e6. Bian ZX, Moher D, Dagenais S, Li YP, Wu TX, Liu L, et al. Improving the quality of randomized controlled trials in Chinese herbal medicine, part IV: applying a revised CONSORT checklist to measure reporting quality. Chin J Integr Med 2006b;4:233e42. Bromet E, Andrade LH, Hwang I, Sampson NA, Alonso J, de Girolamo G, et al. Crossnational epidemiology of DSM-IV major depressive episode. BMC Med 2011;9:90. Butler L, Pilkington K. Chinese herbal medicine and depression: the research evidence. Evid Based Complement Altern Med 2013;2013:739716. Chen J, Zhang Z. Jin gui yao lüe. Beijing: People's Medical Publishing House; 2000 [In Chinese]. Chen L, Zhang YM. Clinical observation on treating mild or moderate depression after stroke with Yiqi Huoxue decoction in 60 patients. Tianjing J Tradit Chin Med 2010;27:378e80 [In Chinese]. Chen CY, Chen YF, Tsai HY. What is the effective component in suanzaoren decoction for curing insomnia? Discovery by virtual screening and molecular dynamic simulation. J Biomol Struct Dyn 2008;26:57e64. Chinese Pharmacopoeia Commission. Pharmacopoeia of the People's Republic of China 2010 edition. Beijing. China: Chinese Medical Science Press; 2010. Chinese Psychiatric Society. The Chinese classification of mental disorders. Shandong: Shandong Publishing House of Science and Technology; 2001. Feng LQ. Pharmacological effects of Western red flower and the identification of its common adulterants. Cap Med 2014;6:48e9 [In Chinese]. Ferrari AJ, Charlson FJ, Norman RE, Patten SB, Freedman G, Murray CJ, et al. Burden of depressive disorders by country, sex, age, and year: findings from the global burden of disease study 2010. PLOS Med 2013;10:e1001547. Fu JL, Zhao YW, Sun XJ. Efficacy and safety of deanxit combined with Wuling capsule in treating post-stroke depression: a randomized controlled trial. J Chin Integr Med 2008;6:258e61 [In Chinese]. Fugh-Berman A. Herbal supplements: indications, clinical concerns, and safety. Nutr Today 2002;37:122e4. Guo ZK, Kong FX, Wu JP. Efficacy of Chinese herbal medicine using “three-phase, five-organ, ten-syndrome differentiation” in treating acute depression. J Shaanxi Coll Tradit Chin Med 2012;35:35e9 [In Chinese]. Higgins JPT, Green S, editors. Cochrane handbook for systematic reviews of interventions version 5.1.0; 2011. Hsu MC, Creedy D, Moyle W, Venturato L, Tsay SL, Ouyang WC. Use of complementary and alternative medicine among adult patients for depression in Taiwan. J Affect Disord 2008;111:360e5. Hu Y. The study of macrophageerelated factors associated with behaviors in depressed mice and the regulation of acupuncture and TCM treatment. Nanjing: Nanjing University of Chinese Medicine; 2011 [Thesis, in Chinese]. Hu SY, Li YH, Zhang CH, Wang SE, Xiao GL. Effects of Chaihu Shugan San and its separated recipes on behavior and CRH、ACTH in plasma of chronic mild unpredicted stress depression rats. Tenth Conference of China Association of Integrative Medicine Professional Committee of mental illness; 2010. Izzo AA, Ernst E. Interactions between herbal medicines and prescribed drugs: an updated systematic review. Drugs 2009;69:1777e98. Jadad AR, Moore RA, Carroll D, Jenkinson C, Reynolds DJ, Gavaghan DJ, et al. Assessing the quality of reports of randomized clinical trials: is blinding necessary? Control Clin Trials 1996;17:1e12. Kessler RC, Soukup J, Davis RB, Foster DF, Wilkey SA, Van Rompay MI, et al. The use of complementary and alternative therapies to treat anxiety and depression in the United States. Am J Psychiatry 2001;158:289e94. Kou MJ, Chen JX. Integrated traditional and Western medicine for treatment of depression based on syndrome differentiation: a meta-analysis of randomized controlled trials based on the Hamilton depression scale. J Tradit Chin Med 2012;32:1e5. Leung KS, Bian ZX, Moher D, Dagenais S, Li YP, Liu L, et al. Improving the quality of randomized controlled trials in Chinese herbal medicine, part III: quality control of Chinese herbal medicine used in randomized controlled trials. Chin J Integr Med 2006;4:225e32. Li YJ, Luo HC, Qian RQ, Zhao XY, Xin HL, Bi J. Effect of Danzhi Xiaoyao powder on neuro-immuno-endocrine system in patients with depression. Chin J Integr Tradit West Med 2007;27:197e200 [In Chinese]. Liao JF, Jan YM, Huang SY, Wang HH, Yu LL, Chen CF. Evaluation with receptor binding assay on the water extracts of ten CNS-active Chinese herbal drugs. Proceeding Natl Sci Counc Repub China Part B 1995;19:151e8.

175

Lin H, Ye X, Guo ZP, Liao YL. Combination of paroxetine and Xiaoyaosan for postpartum depression. Appl J General Pract 2008;6:476e7 [In Chinese]. Luo HC, Liu P, Meng FQ, Chen YC, Zhao XY, Tan CX, et al. Combined utilization of amitriptyline and Ginkgo Biloba extract in the treatment of depression e a multicenter randomized double blind comparison. Chin Ment Health J 1999;13: 167e9 [In Chinese]. Manchikanti L, Ailinani H, Koyyalagunta D, Datta S, Singh V, Eriator I, et al. A systematic review of randomized trials of long-term opioid management for chronic non-cancer pain. Pain Physician 2011;14:91e121. Ministry of Health of the People's Republic of China. Pharmacopoeia of the People's Republic of China. Shanghai: China Medical Science and Technology Press; 2010. Mizowaki M, Toriizuka K, Hanawa T. Anxiolytic effect of Kami-Shoyo-San (TJ-24) in mice: possible mediation of neurosteroid synthesis. Life Sci 2001;69:2167e77. Moshiri E, Basti AA, Noorbala AA, Jamshidi AH, Abbasi SH, Akhondzadeh S. Crocus Sativus L. (petal) in the treatment of mild-to-moderate depression: a doubleblind, randomized and placebo-controlled trial. Phytomedicine 2006;13:607e11. Noorbalaa AA, Akhondzadeha S, Tahmacebi-Poura N, Jamshidib AH. Hydro-alcoholic extract of Crocus Sativus L. versus fluoxetine in the treatment of mild to moderate depression: a double-blind, randomized pilot trial. J Ethnopharmacol 2005;97:281e4. Qin XK, Li P, Han M, Liu ZJ, Liu JP. Systematic review of randomize controlled trials of Xiaoyao Powder in treatment of depression. J Tradit Chin Med 2010;51:500e5. Qin F, Wu XA, Tang Y, Huang Q, Zhang ZJ, Yuan JH. Meta-analysis of randomized controlled trials to assess the effectiveness and safety of free and easy Wanderer plus, a polyherbal preparation for depressive disorders. J Psychiatr Res 2011;45:1518e24. Qureshi NA, AI-Bedah AM. Mood disorders and complementary and alternative medicine: a literature review. Neuropsychiatr Dis Treat 2013;9:639e58.  n C, Alcaraz F, Verpoorte R, Heinrich M. What is in a name? Rivera D, Allkin R, Obo the need for accurate scientific nomenclature for plants. J Ethnopharmacol 2013;152:393e402. Sarris J. Chinese herbal medicine for sleep disorders: poor methodology restricts any clear conclusion. Sleep Med Rev 2012;16:493e5. Shen ZM, Zhu ML, Zhao AQ, Yu ZJ, Wang LP, Jiang T, et al. Comparative observation on efficacy of Jieyu pill and maprotiline in treating depression. Chin J Integr Tradit West Med 2004;24:415e7 [In Chinese]. Spitzer RL, Endicott J, Robins E. Research Diagnostic Criteria (RDC) for a selected group of functional disorders. 3rd ed. Research Assessment and Training Unit, New York: New York State Psychiatric Institute; 1989. Sun XY, Chen AQ, Xu XF, Zhang HG, Tang QS, Zhang HY. Randomized double blind, placebo-controlled trial of Shuganjieyu capsule in the treatment of mild or moderate depression. Chin J New Drugs 2009;18:413e7 [In Chinese]. Sun ZG, Huang QZ, Feng J, Chen LP. Clinical study of Shuyusan in treatment of depression. Chin J Info TCM 2012;19:7e10 [In Chinese]. Ta G, Wang J. Clinical observation on treating post-stroke depression with Chaihujialonggu Muli decoction in 24 patients. Jilin J Tradit Chin Med 2008;28: 179e80 [In Chinese]. Tao JQ. Jieyu pill versus venlafaxine in treatment of depression at different levels of severity. Chin J Clin Rehabil 2006;10:1e3 [In Chinese]. Tong Y, Zou J, Liu P, Ni LQ, Cao BY, Liang SH, et al. The effect of four kinds of traditional Chinese herbal compound on the experimental acute stress behaviors and the hypothalamus-pituitary-adrenal gland axis. Zhongguo Zhong Yao Za Zhi 2005;30:1863e6 [In Chinese]. Wang Y, Fan R, Huang X. Meta-analysis of the clinical effectiveness of traditional Chinese medicine formula Chaihu-Shugan-San in depression. J Ethnopharmacol 2012;141:571e7. White AR, Ernst E. A systematic review of randomized controlled trials of acupuncture for neck pain. Rheumatology 1999;38:143e7. World Health Organization. WHO expert committee on specifications for pharmaceutical preparations. Fortieth report. Geneva: World Health Organization; 2006. World Health Organization. Mental Health Atlas 2011. Geneva: World Health Organization; 2011. Wu LL, Yan C, Su JF, Xu ZW, Wang WZ. Effect of Dan Zhi Xiao Yao San and its extracts on hypothalamus - pituitary gland - adrenal gland secretion in rats with chronic psychological stress. J Guangzhou Univ Traditional Chin Med 2006;23:413e5. Yang RX, Dong Y, Wang DM. Clinical observation of Shujiele Wutang pill in the treatment of depression. Chin Med Focus 2002;5:16e7 [In Chinese]. Yang LQ, Chen JL, Yu LT, Wu XQ, Zhang Y, Xu F, et al. Qican Fukang capsule in the treatment of chronic depression in 445 patients. Shaanxi J Tradit Chin Med 2012;33:1032e5 [In Chinese]. Yi PL, Tsai CH, Chen YC, Chang FC. Gamma-aminobutyric acid (GABA) receptor mediates suanzaorentang, a traditional Chinese herb remedy -induced sleep alteration. J Biomed Sci 2007;14:285e97. Yi ZH, Zhu LP, Long B, Yao PF, Zhao GX, Chen AZ, et al. Clinical observation on treatment of major depression disorder by paroxetine combined with Chaihu Xiaoyao mixture. Chin Med Integr Tradit West Med 2010;30:1257e60 [In Chinese]. Zhang MJ, Zhang QY, Cui GB. The clinical study of Xiaoyao san on neurotic depression. J Shandong Univ Chin Med 1998;22:34e7 [In Chinese]. Zhang Y, Han M, Liu Z, Wang J, He Q, Liu J. Chinese herbal formula Xiao Yao San for treatment of depression: a systematic review of randomized controlled trials. Evidence-Based Complementary and Alternative Medicine 2012; 2012. p. 931636. Zhao H, Wan X, Chen JX. A mini review of traditional Chinese medicine for the treatment of depression in China. Am J Chin Med 2009;37:207e13. Zhao SQ, Zhu L, Zhao P, Rao R. Clinical observation on Ganmai Dazao decoction treatment for depression. Clin J Tradit Chin Med 2012;24:731e2.