You are on page 1of 10

+Model

EUJIM-308; No. of Pages 10 ARTICLE IN PRESS


Available online at www.sciencedirect.com

ScienceDirect
European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10

Original article

Chinese Herbal Medicine for premature ovarian failure: A systematic review


and meta-analysis
Yan Wu a , Lu-Ting Chen a , Fan Qu a , Irfan S. Sheikh b , Yan-Ting Wu a,∗
a Department of Reproductive Endocrinology, Women’s Hospital, School of Medicine, Zhejiang University, Hangzhou, Zhejiang 310006, China
b School of Medicine, Zhejiang University, Hangzhou, Zhejiang, China

Received 31 July 2013; received in revised form 16 January 2014; accepted 17 January 2014

Abstract
Introduction: Premature ovarian failure (POF), loss of normal ovarian function before the age of 40 commonly results in infertility. This systematic
review/meta-analysis evaluates the effectiveness of Chinese herbal treatments on POF.
Methods: Systematic searches of six electronic databases were conducted for articles published through June 2013. All randomized controlled
trials which compared Chinese Herbal Medicine (CHM) as solitary treatment vs. hormone replacement therapy (HRT) for POF were included.
Measurement of treatment effect was done through pooled odds ratio (OR) of effective cases in the two groups, and weighted mean difference
(WMD) of hormonal levels (E2 , FSH, LH) after treatment.
Results: This meta-analysis includes 10 randomized controlled trials (RCTs) involving 888 patients in comparing CHM with HRT for the treatment
of POF. The pooled data showed an improvement in symptoms in the CHM group compared to that of the HRT group (OR = 2.50, 95% CI: 1.49–4.18,
P < 0.001). FSH levels were significantly lower in patients treated with CHM compared to that of patients treated with HRT (≥6 M: WMD = −8.34,
95% CI: −11.96, P < 0.001, −4.71; <6 M: WMD = −5.23, 95% CI: −8.68, −1.78, P = 0.003). There was no significant difference (P > 0.05)
observed in E2 and LH levels when CHM and HRT were compared.
Conclusions: CHM may relieve symptoms of POF partly through decreasing serum FSH levels, this may provide guidance for future studies.
© 2014 Elsevier GmbH. All rights reserved.

Keywords: Premature ovarian failure; Chinese Herbal Medicine; Hormone replacement therapy

Introduction vaginitis, and mood swings [4]. In many cases, patients suf-
fer from infertility due to the absence of follicles and failure to
Premature ovarian failure (POF) is the loss of ovarian func- respond to stimulation of hormones [5].
tion which is characterized by amenorrhea, hypoestrogenism, A definitive diagnostic criterion for POF has not been estab-
and elevated serum gonadotropin levels in women younger than lished [6]. In clinical practice, a diagnosis is usually based on
the age of 40. The estimated incidence of POF is 1:10000 women amenorrhea for 3–6 months, elevated serum follicle-stimulating
by age 20; 1:1000 women by age 30; and 1:100 women by age 40 hormone (FSH) levels (usually above 40 IU/L) detected on at
[1,2]. This condition occurs in 10–28% of women with primary least two separate occasions several months apart, and low estro-
amenorrhea and 4–18% with of women with secondary amen- gen levels [7]. Other causes of amenorrhea such as pregnancy,
orrhea [3]. Women with POF experience menopausal symptoms polycystic ovarian syndrome, and thyroid dysfunction need to
such as hot flashes, vaginal dryness, dyspareunia, insomnia, be excluded. Although specific standardized guidelines for the
management of POF are absent, hormone replacement ther-
apy (HRT) with estrogen or progesterone is widely considered
as appropriate medical options [6]. However the use of HRT
∗ Corresponding author at: Department of Reproductive Endocrinology, has been associated with increased risks of breast cancer, heart
Women’s Hospital, School of Medicine, Zhejiang University, 1 Xueshi Road,
Hangzhou, Zhejiang 310006, China. Tel.: +86 57187061501;
attacks and strokes [8].
fax: +86 57187061878. For thousands of years, Chinese Herbal Medicine (CHM) has
E-mail address: yanting wu@163.com (Y.-T. Wu). been used to treat many diseases. The philosophy of Chinese

1876-3820/$ – see front matter © 2014 Elsevier GmbH. All rights reserved.
http://dx.doi.org/10.1016/j.eujim.2014.01.008

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
xxx.e2 Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10

Medicine in terms of biology, pathophysiology and pharma- hormone levels of E2 , FSH and LH after treatment in both CHM
cology takes an entirely different approach. When compared groups and HRT groups. The number of events, participants
to Western Medicine, the diagnostic and therapeutic modalities in each group, mean and standard deviation for hormone levels
rely on experience and social customs. For this reason, many were clearly defined. Case reports, reviews, animal experiments,
treatments still do not have a concise mechanism of action and non-randomized controlled trials, and studies failing to mention
clear side effect profile. Currently, CHM are mainly used to Chinese herbs or using CHM plus other treatments compared
decrease the risk and treat common pathologies i.e. hyperten- with solitary treatment with HRT were excluded.
sion, hypercholesterolemia, diabetes mellitus, obesity, cancer,
depression, prevention of miscarriages and even POF. People Search methods for identification of studies
mainly use CHM either because they experience fewer side
effects and/or they have had no effect from Western Medicine We searched the Cochrane Menstrual Disorders and Subfer-
treatments. The outlook that CHM takes on POF is a unique tility Group trials register (June 2013), MEDLINE (1966 – June
one. Animal experiments have demonstrated the efficacy of Chi- 2013), EMBASE (1974 – June 2013), China National Knowl-
nese herbs to inhibit the production of anti-ovarian antibodies edge Infrastructure (CNKI, 1982 – June 2013), Wanfang Data
(AoAb) and to further protect the ovaries from autoimmune (1982 – June 2013) and the Chinese Scientific and Technical
destruction [9]. Studies have been done to reveal CHM’s influ- Journals database (VIP, 1989 – June 2013). No restrictions of
ence on the serum level of hormones and its involvement with language or publication type were placed in any of the searches.
the hypothalamic–pituitary–ovarian (HPO) axis [10,11]. For The following free text terms were searched: Traditional Chi-
example, several case reports suggest that CHM may alleviate nese Medicine, Chinese Herbal Medicine, Herbal Drugs, Plant
symptoms, reduce elevations in FSH, promote estrogen produc- Extracts, Decoction, Tang. Medical subject headings were also
tion, and regulate levels of luteinizing hormone (LH) [12,13]. searched, which included the terms: Recipe and premature ovar-
CHM is also effective in the management of the infertility caused ian failure, primary ovarian insufficiency, premature menopause,
by POF [14]. However, evidence in terms of larger sample sizes premature ovarian dysfunction, amenorrhea, and menstruation
and multi-center, randomized controlled trials (RCTs) are lack- disturbance.
ing. In addition, when comparing the outcomes of CHM and
HRT in terms of treatment profile (effective cases) and overall Selection of studies
efficacy (hormone levels) the results are controversial. We herein
provide a systematic review and meta-analysis of the available A total of 379 citations were obtained from electronic search
literature to compare the outcomes of CHM and HRT in POF engines. After screening the titles and abstracts, 319 articles that
treatment. We measure the effectiveness of treatment based on did not fit the eligibility criteria were excluded. Further exclu-
primary outcome (relief of symptoms of POF) and secondary sion of studies was taken after review of the selected studies.
outcome (serum hormone levels). Studies with incomplete data of all hormonal levels (E2 , FSH,
LH) and unclear number of valid cases, or that did not meet
Materials and methods our requirements of group stratification were excluded. At last,
10 published RCTs were selected for meta-analysis after strict
Eligibility criteria exclusion and inclusion criteria were implemented [12,16–24].
This methodology of exclusion and inclusion criteria is depicted
All RCTs comparing the efficacy of Chinese herbs with HRT by Fig. 1.
in the treatment of POF were included. The diagnostic cri-
teria for POF included women who were age 40 and below Assessments bias risk
who presented with amenorrhea for more than four months
with perimenopausal symptoms, such as chills and pain of the Two independent reviewers assessed the risk of bias with
waist and the knees, vaginal dryness, and sexual hypoactivity. the criteria in the Cochrane Handbook for Systematic Reviews
These symptoms along with ancillary laboratory findings such of Interventions 5.1.0 [15]. Sequence generation, allocation
as FSH levels greater than 40 IU/L, and estrogen (E2 ) less than concealment, blinding, incomplete data assessment, selective
110 mol/L were enough to diagnose patients with POF. Other outcome reporting and other potential sources of bias were
diseases such as insensitive ovarian syndrome and gonadal dys- assessed. Any disagreements between reviewers were resolved
genesis were excluded. Baseline characteristics, such as age and by discussion and arbitration by a third reviewer (FQ).
course, in the patients of the two groups were required to be
comparable. The therapeutic intervention was any CHM with Data extraction and management
the function of tonifying the kidney, including decoction, pill,
granule and recipe. The outcome measures included primary Data were carefully extracted from all eligible articles inde-
and secondary measurements. Our primary measurement was pendently by two reviewers (YW and YTW). Relevant data
the effective cases which comprised of relief of symptoms of that was collected included the year of publication, country of
POF including the recurrence of menstruation, improvement study, study design, patient characteristics, sample size, and out-
of chills and pain (the waist and knees), alleviation of vaginal come measurements. Treatment effect was measured through the
dryness, and increased libido. Our secondary measurement was pooled odds ratio (OR) for significant efficiency and differences

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10 xxx.e3

Fig. 1. The process of study selection for the systematic review with meta-analysis of the efficacy of Chinese Herbal Medicine in the treatment of premature ovarian
failure.

in hormone levels (E2 , FSH, LH) after treatment with CHM and of tonifying the kidney and HRT’s. Details involving the dose,
HRT. Safety was also an important outcome, however there were frequency, and duration are presented in Table 3. All of these
no unified quantitative indicators for it; we therefore provide a 10 studies were conducted in China and were single-center ran-
brief description of safety issues that were noted. domized controlled trials. Six of them found the effective rate
to be significantly higher in the CHM group, and others showed
Statistical analysis no significant difference between CHM and HRT groups.

Statistical analysis was performed with STATA 12.0 software Risk of bias in included studies
(StataCorp, College Station, USA). The results were pooled and
expressed as OR with 95% confidence interval (CI) for count Risk for bias was measured for each of the ten included stud-
data and weighted mean difference (WMD) with 95% CI for ies. For most of the studies, risk of bias was high or unclear for
continuous data. Statistical heterogeneity was tested using the many of the domains we had assessed. The reports of all trials
Chi-squared test and I2 statistic. We used the fixed-effects model mentioned randomization, but only five studies [12,17,18,22,23]
when there was pool heterogeneity (I2 < 50%, p > 0.05); other- described the specific randomization strategies. Two of the five
wise, the random-effects model was adopted. Subgroup analyses studies [12,23] grouped patients according to their registration
were performed according to the duration of treatment (≥6 order, which might have led to a high risk for selection bias.
months or <6 months). Sensitivity analysis was done to examine Three of the ten studies had a low risk for selection bias as they all
the effect of excluding each study. In the meta-analysis, we used had adequate methodology for randomization by random num-
Forest Plots in assess the heterogeneity of treatment effects and ber table method [17,18,22]. We also graded the included studies
Funnel Plots in the evaluation of publication bias. in respect to other domains, such as allocation concealment,
blinding of participants and personnel, blinding of outcome
Results assessment, incomplete outcome data, and selective reporting.
These domains were graded according to the guidelines provided
Main study characteristics by the Cochrane Handbook 5.1.0. All of the included studies had
a low risk for attrition bias which was due to our strict eligibility
After the critical process of study selection, 10 published criteria needed to perform our meta-analysis. Table 4 shows the
RCTs were obtained for meta-analysis after exclusion and risk of bias assessed in the 10 included studies comparing CHM
inclusion criteria were implemented [12,16–24]. Baseline char- vs. HRT in the treatment of POF.
acteristics and overall quality of the 10 studies were recorded
and provided in tabular format (Tables 1 and 2). This study evalu- Effects of interventions
ated 888 patients with POF: 466 patients were treated with CHM
and 422 patients were treated with HRT. The patients were all All 10 included RCTs studied the clinical effects and hor-
younger than 40 with disease length of at least four months. The monal levels after treatment of CHM vs. HRT. We first studied
interventions included different kinds of CHM with the function the differences of effective cases between CHM and HRT

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
xxx.e4 Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10

Fig. 2. Forest plot of effect sizes for Chinese Herbal Medicine in premature ovarian failure treatment.

groups. The pooled data showed significant efficacy (P < 0.001) the CHM group. Additionally, there were no significant differ-
of CHM in the treatment of POF (OR = 2.50, 95% CI: 1.49–4.18, ences in LH levels (WMD = −0.86, 95% CI: −2.52 to 0.79,
N = 888) (Fig. 2). In addition, publication bias was analyzed P = 0.308) in these two groups. Begg’s and Egger’s test showed
by Begg’s funnel plot (Fig. 3), which did not reveal any sig- there were no significant (P > 0.05) publication biases (Table 5).
nificant evidence of asymmetry. Secondly, we compared the
hormone levels after treatment with CHM and HRT, and found Sensitivity analysis
that E2 levels were significantly higher (WMD = 22.00, 95%
CI: 2.90–41.10, P = 0.024) and FSH levels were significantly Sensitivity analysis should been done when the statistical het-
lower (WMD = −6.59, 95% CI: −9.06 to −4.12, P < 0.001) in erogeneity is high. It was performed to investigate the influence

Table 1
Interventions for Chinese herbs and hormone replacement therapy used in 10 randomized trials done for the treatment of premature ovarian failure.
Study Number of patients Intervention Effective cases
(CHM/HRTa ) (CHM/HRT)

Gao H et al. (2007) 30 vs. 30 Bushen Tiaochong Recipe vs. Conjugated 28 vs. 27
Estrogens + Medroxyprogesterone
Lu XN et al. (2008) 40 vs. 20 Bushen Er’xian decoction vs. Conjugated 29 vs. 9
Estrogens + Medroxyprogesterone
Han LY et al. (2008) 36 vs. 58 Kangluanshuai granules vs. Diethylstibestrol + Medroxyprogesterone 26 vs. 41
Cao HN (2009) 36 vs. 24 Er’zhi Dihuang decoction vs. Estradio + Medroxyprogesterone 29 vs. 19
Ning Y et al. (2010) 30 vs. 30 Bushen Yichong pill vs. Estradio + Medroxyprogesterone 26 vs. 27
Zhao RX (2010) 36 vs. 30 Jiawei Er’xian decoction vs. Conjugated Estrogens + Medroxyprogesterone 31 vs. 16
Ding Q et al. (2010) 30 vs. 30 Huluan decoction vs. Desogestrel + Ethinylestradiol 26 vs. 21
Liang Z et al. (2010) 42 vs. 38 Kangshuai Fuchao decoction vs. Conjugated 37 vs. 31
Estrogens + Medroxyprogesterone
Zhu ZZ et al. (2011) 51 vs. 32 Heche Zibu decoction vs. Conjugated Estrogens + Medroxyprogesterone 47 vs. 14
Yang XT (2011) 135 vs. 130 Bushen decoction vs. Conjugated Estrogens + Medroxyprogesterone 98 vs. 58
a CHM: Chinese Herbal Medicine; HRT: hormone replacement therapy.

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10 xxx.e5

Table 2
Quality assessment of 10 randomized trials done to evaluate Chinese Herbs vs. hormone replacement therapy in the treatment premature ovarian failure.
Study Country Single-/multi center Randomization method Concealment of allocation Blinding

Gao H et al. (2007) China Single-center Random number table method Unclear No
Lu XN et al. (2008) China Single-center Registration order Unclear No
Han LY et al. (2008) China Single-center Not mentioned Unclear No
Cao HN (2009) China Single-center Not mentioned Unclear No
Ning Y et al. (2010) China Single-center Not mentioned Unclear No
Zhao RX (2010) China Single-center Registration order Unclear No
Ding Q et al. (2010) China Single-center Random number table method Unclear No
Liang Z et al. (2010) China Single-center Not mentioned Unclear No
Zhu ZZ et al. (2011) China Single-center Not mentioned Unclear No
Yang XT (2011) China Single-center Random number table method Unclear No

Table 3
Details of CHM Interventions including Dosing, Usage, and Duration.
CHM Dosing Usage Duration

Bushen Tiaochong Recipe Semen Cuscutae 15 g, Rhizoma Polygonati 15 g, Radix Add 300 ml of water to decoction. 3 months
Rehmanniae Preparata 15 g, Herba Cistanches 10 g, Radix Take 150 ml every 12 h.
Morindae Officinalis 10 g, Radix Angelicae Sinensis 10 g,
Rhizoma Ligustici 6 g, Fluoritum 15 g, Fructus Schisandrae
Chinensis 6 g.
Bushen Er’xian Decoction Radix Rehmanniae Preparata 20 g, Rhizoma Dioscoreae 9 g, Add 300 ml of water to decoction. 6 months
Fructus Corni 9 g, Cortex Cinnamomi 9 g, Rhizoma Take 150 ml every 12 h.
Curculigins 9 g, Epimedium 9 g, Radix Angelicae Sinensis
9 g, Radix Glycyrrhizae 3 g.
Kangluanshuai granules Radix Rehmanniae Preparata, Rhizoma Dioscoreae, Fructus Take 27 g daily, 9 g TID. 3 months
Corni, Poria, Rhizoma Alismatis, Moutan, Fructus Ligustri
Lucidi, Herba Ecliptae and so on.
Er’zhi Dihuang Decoction Fructus Ligustri Lucidi 30 g, Eclipta 30 g, Radix Rehmaniae Add 300 ml of water to decoction. 3 months
Recens 30 g, Semen Cuscutae 10 g, Fructus Lycii 10 g, Take 150 ml every 12 h.
Rhizoma Polygonati 10 g, Rhizoma Ligustici 10 g, Flos
Carthami 10 g, Placenta Hominis 3 g, Radix Angelicae
Sinensis 15 g.
Bushen Yichong Pills Radix Rehmanniae Preparata, Radix Angelicae Sinensis, Take 30 g daily, 10 g TID from day 5 9 months
Radix Paeoniae Alba, Fructus Corni, Rhizoma Polygonati, to day 26 in a menstrual cycle.
Placenta Hominis and so on.
Jiawei Er’xian Decoction Radix Astragali seu Hedysari 30 g, Radix Codonopsis 30 g, Add 300 ml of water to decoction. 6 months
Radix Rehmanniae Preparata 15 g, Radix Angelicae Take 150 ml every 12 h.
Sinensis 15 g, Rhizoma Curculigins 10 g, Herba Epimedii
10 g, Radix Morindae Officinalis 10 g, Fructus Corni 10 g,
Rhizoma Dioscoreae 10 g, Cortex Phellodendri 3∼6 g,
Rhizoma Anemarrhenae 3∼6 g, Cortex Cinnamomi 6 g,
Radix Glycyrrhizae 6 g.
Huluan Decoction Fluoritum 15 g, Semen Cuscutae 10 g, Herba Cistanches Add 300 ml of water to decoction. 3 months
10 g, Fructus Psoraleae 10 g, Fructus Rubi 10 g, Radix Take 150 ml each time, 30 min after
Linderae 10 g, Herba Dendrobii 10 g, Fructus Mori 10 g, breakfast and dinner.
Lotus plumule 6 g, Radix Glycyrrhizae 6 g.
Kangshuai Fuchao Decoction Radix Codonopsis 20 g, Radix Astragali seu Hedysari 30 g, Add 300 ml of water to decoction. 9 months
Fructus Lycii 20 g, Ttortoise Plastron 15 g, Semen Cuscutae Take 150 ml every 12 h.
20 g, Herba Cistanches 20 g, Cortex Moutan Radicis 20 g,
Ramulus Cinnamomi 10 g, Radix Paeoniae Alba 15 g,
Radix Angelicae Sinensis 15 g, Mix-fried Licorice 10 g.
Heche Zibu Decoction Placenta Hominis 20 g, Radix Rehmanniae Preparata 20 g, Add 300 ml of water to decoction. 4 months
Carapax et Plastrum Testudinis 20 g, Radix Achyranthis Take 100 ml every 8 h.
Bidentatae 15 g, Cortex Phellodendri 15 g, Cortex
Eucommiae 15 g, Radix Ginseng 10 g, Radix Asparagi 10 g,
Radix Ophiopogonis 10 g.
Bushen Decoction Radix Rehmanniae Preparata 15 g, Rhizoma Dioscoreae Add 300 ml of water to decoction. 6 months
10 g, Cornus 10 g, Rhizoma Alismatis 9 g, Poria 9 g, Cortex Take 150 ml every 12 h.
Cinnamomi 6 g, Radix Morindae Officinalis g, Fructus
Ligustri Lucidi 1 g, Radix Glehniae 12 g, Radix Astragali
seu Hedysari 30 g, Radix Angelicae Sinensis 12 g, Radix
Codonopsis 30 g, Radix Glycyrrhizae 6 g.

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
xxx.e6 Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10

Table 4
Risk of bias assessed in 10 randomized studies comparing CHM vs. HRT in the treatment of premature ovarian failure.
Study Random sequence Allocation Blinding of participants Blinding of outcome Incomplete Selective
generation concealment and personnel assessment (detection outcome data reporting
(selection bias) (selection bias) (performance bias) bias) (attrition bias) (reporting bias)

Gao H et al. (2007) Low Unclear High High Low Unclear


Lu XN et al. (2008) High Unclear High High Low Unclear
Han LY et al. (2008) Unclear Unclear High High Low Unclear
Cao HN (2009) Unclear Unclear High High Low Unclear
Ning Y et al. (2010) Unclear Unclear High High Low Unclear
Zhao RX (2010) High Unclear High High Low Unclear
Ding Q et al. (2010) Low Unclear High High Low Unclear
Liang Z et al. (2010) Unclear Unclear High High Low Unclear
Zhu ZZ et al. (2011) Unclear Unclear High High Low Unclear
Yang XT (2011) Low Unclear High High Low Unclear

Table 5
Evaluation of publication bias.
No. of studies No. of cases Begg’s test (P) Egger’s test (P)

Effective cases (CHM/HRT) 10 888 0.474 0.542


E2 (CHM vs. HRT) 10 888 0.210 0.112
FSH (CHM vs. HRT) 10 888 0.858 0.469
LH (CHM vs. HRT) 10 888 0.858 0.490

of a single study on the overall meta-analysis estimate. In our in subgroup analyses were similar to that of all 10 studies
study, sensitivity analyses of effective cases and hormonal levels (Figs. 4 and 5). However, there was no significant difference
were carried out by systematically excluding studies one-by-one of E2 levels between CHM and HRT in both of the subgroups
and analyzing the homogeneity and effect size of the remaining (Fig. 6). The results may suggest CHM to have similar curative
studies. As a result, no single study qualitatively changed the properties as HRT in the treatment of POF. Furthermore, we
pooled OR of effective cases and WMD of hormonal levels, report the heterogeneity of the long-course treatment group (≥6
indicating the stability of the results of the meta-analysis. month) was smaller than the short-course treatment group (<6
months).
Subgroup analyses of hormone levels
The safety of CHM in treating POF
To explore the effect of duration of CHM, we conducted sub-
group analyses. We divided these 10 studies into two groups Routine blood count, urinalysis, stool, liver function and renal
based on the length of treatment (≥6 months and <6 months) function tests, EKG, breast ultrasound transvaginal ultrasound
and placed them into tabular format (Table 6). The results of were used in some studies to follow up on any adverse effects
comparisons of FSH and LH levels between CHM and HRT during the course of treatment. Of all ten studies, two studies
[16,20] did not assess adverse events; four studies [12,17,23,24]
described non adverse reactions in both CHM and HRT groups.
Four studies reported adverse events, including gastrointesti-
nal discomfort, breast pain, vaginal bleeding, headache and
edema [18,19,21,22]. Among these four studies, two studies
[18,19] emphasized these side effects occurring at a lower rate
in the CHM group compared with the HRT group, but no severe
adverse drug reactions were found in the afore mentioned two
groups.

Discussion

Summary of main results

According to its clinical features, POF is attributed to “Kid-


ney deficiencies” in traditional Chinese medicine. “Kidney
Fig. 3. Funnel plot of effect sizes for Chinese Herbal Medicine in premature tonics” are therapeutic herbal treatments that are used against
ovarian failure treatment. “Kidney deficiencies”. However, the therapeutic value has not

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10 xxx.e7

Table 6
Subgroup analyses of hormone levels post-treatment with CHM and HRT.
Subgroups No. of studies No. of cases Summary Effect Study Heterogeneity

WMD (95% CI) p Value I2 , % p Value

≥6 M 5 531 6.61 (−1.24, 14.46) 0.099 55.5 0.061


E2 (CHM vs. HRT)
<6 M 5 357 37.70 (−2.77, 78.18) 0.068 98.9 0.000
≥6 M 5 531 −8.34 (−11.96, −4.71) 0.000 52.1 0.079
FSH (CHM vs. HRT)
<6 M 5 357 −5.23 (−8.68, −1.78) 0.003 80.7 0.000
≥6 M 5 531 0.40 (−1.32, 2.11) 0.650 0.0 0.941
LH (CHM vs. HRT)
<6 M 5 357 −1.82 (−4.45, 0.82) 0.176 68.9 0.012

been extensively studied in terms of clinical trials. This sys- Probable sources of heterogeneity
tematic review shows a trend for Chinese herbs to improve
symptoms of POF as seen from the effective cases as well as In these 10 included studies, no two had the same treatment
improvement in hormone levels by increasing E2 , and decreas- options that may have led to great heterogeneity. In addition,
ing FSH and LH levels. CHM was even better than HRT in some the duration of treatment in each of the studies also varied
areas of comparisons, such as effective cases and decreased lev- potentially affecting the efficiency of the Chinese herbs. Our
els of FSH. In addition, the heterogeneity was smaller in the subgroup analyses showed a smaller heterogeneity in long-
long-course treatment subgroup (≥6 months), which may indi- course treatment when compared with short-course treatment.
cate that long term treatment have more stable results when The sample sizes in some studies were small, which might lead
compared with short term treatment. to wide variances in weight distributions and might be a source

Fig. 4. Forest plot of E2 levels after treatment with CHM and HRT.

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
xxx.e8 Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10

Fig. 5. Forest plot of FSH levels after treatment with CHM and HRT.

of heterogeneity. The specific measurements of hormones and studies as there was no mention of patients who were loss to
the time as to when the data was collected were not mentioned follow-up and patients were not adequately randomized, respec-
in some studies. Other affecting factors include the quality of tively. Finally, original data of the studies was not available,
the evidence as mentioned above. further limiting thorough evaluation of the efficiency of Chinese
herbs for POF.
Quality of evidence and potential biases
The safety of CHM in treating POF
The description of the study design and accuracy of the data
are critical in conducting a meta-analysis. The method of blind- Drug safety is extremely important and should be commented
ing had failed to be mentioned in all 10 studies selected for upon in each study. Possible adverse effects of HRT included an
meta-analysis, which may have led to selection and detection increased risk of breast cancer, heart attacks and strokes [8].
biases. The randomization strategies were inadequate in 7 of It is a popular belief that the side effects of Chinese herbs are
these 10 studies [12,16,19–21,23,24], which could have led to much milder when compared to Western medicine. However,
selection bias. In addition, the treatment options including the Chinese herbs have complicated constituents, and it is difficult
composition of regimen and dose of CHM were varied in these to define their side effects through pharmacologic analysis. Fur-
10 included studies, which may have led to the great heterogene- ther details encompassing these 10 studies are discussed in the
ity observed in the results. In these included 10 studies, there was above section. We therefore conclude from these parameters that
only one had a sample size larger than 100 [22], which may have Chinese herbs are safe for most patients with POF, however this
led to wide variances in weight distributions and may have been statement is generated based upon the unequivocal amount of
another source of heterogeneity. data published on the side effects of CHM.
Some limiting factors that were met included the following
parameters. First, the provided analysis is only based on pub- Potential mechanism of CHM action
lished data and no unpublished data were included. Second, all
included studies are single-center studies done in China. Further- Besides chemotherapy and radiation treatments for cancer,
more, recall or selection bias may be confronted in the included the causes of POF are usually idiopathic. Some causes of POF

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10 xxx.e9

Fig. 6. Forest plot of LH levels after treatment with CHM and HRT.

include genetic disorders (i.e. Turner Syndrome and Fragile X And long term treatment is better than short term treatment when
Syndrome) and autoimmune disorders (i.e. Hashimoto’s Thy- comparing heterogeneity. However, in the process of study selec-
roiditis and Addison’s disease). An autoimmune mechanism was tion, we strongly advise for major improvements in adequate
first suggested as a possible etiology in 1966 when antibodies to blinding techniques, randomization method and simple size, if
the cytoplasm of rabbit ova were found by Vallolton and Forbes one is to thoroughly evaluate the beneficial effects of CHM. In
[25]. Two of the 10 included RCTs in the included studies also addition, the regimens of CHM in treating POF vary but if we
compared negative transformation rate of anti-ovarian antibody can extract the main active components to standardize the treat-
[16,21]. Both of these studies found the transformation rate to be ment dose, there may be opportunities to study the efficiency of
higher in the CHM group. T lymphocytes play an important role single herbs.
in adaptive immunity, and deficiencies in estrogen may affect the To date there are no studies that compare Chinese Herbal
apoptosis of T lymphocytes. Lu XN et al. found an increase of Medicine with hormone replacement therapy in the management
CD4 and a decrease in CD8T Lymphocytes after treatment with of patients suffering from premature ovarian failure. Chinese
CHM compared with HRT [12]. Further suggesting that CHM Herbal Medicines are not well known to most Western physi-
has inhibitory actions in autoimmune processes that regulate T cians. This review provides an introduction to some Chinese
cell immunity, however, the specific mechanisms are not well herbs which have minimal side effects and which could poten-
defined. tially help physicians manage their patients suffering from POF.
The current literature provides some evidence to support the use
of CHM in clinical practice. The treatment modalities, record-
Implications for further studies and clinical applications ing of any side effects, and defining the treatment algorithm
of CHM drugs needs further investigation. Until then there is
We found that Chinese herbs substantially improve POF need for vigilance for any abnormal findings reported by patients
based on the risk of bias, and lack of detail reported on pri- receiving such treatments. Physicians should use their best judg-
mary studies. Hormone levels also improved after treatment ment when determining the time to follow up after usage of
with CHM when compared to hormone levels prior to treatment. CHM.

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008
+Model
EUJIM-308; No. of Pages 10 ARTICLE IN PRESS
xxx.e10 Y. Wu et al. / European Journal of Integrative Medicine xxx (2014) xxx.e1–xxx.e10

Conclusion ovarian failure [in Chinese]. Traditional Chinese Drug Research & Clinical
Pharmacology 2012;23:381–6.
[11] Zhang DW, Li CP, Huang X, Zhang DD, Fan YL, Zhang Y, et al. Effect
From our meta-analysis, CHM may relieve symptoms of POF
of invigorating kidney and regulating periphery on rats with premature
partly through decreasing serum FSH levels, which may provide ovarian failure and its influence on inhibin B and wascular endothelial
a guide for future study design in the field. growth factor [in Chinese]. Chinese Journal of Experimental Traditional
Medical Formulae 2011;17:213–6.
Conflict of interest [12] Lu XN, Xu XR, Lin LJ. Clinical observation of bushen er’xian decoction
in treating premature ovarian failure [in Chinese]. Zhongguo Zhong Xi Yi
Jie He Za Zhi 2008;28:594–6.
No competing financial interests exist. [13] Hua FF, Xia YH, Yang J. Clinical observation on treatment of premature
ovarian failure patients of shen deficiency gan stagnation syndrome by com-
Acknowledgments bination of bushui roumu recipe and medroxyprogesterone acetate tablet.
Zhongguo Zhong Xi Yi Jie He Za Zhi 2012;32:1028–31.
[14] Chao SL, Huang LW, Yen HR. Pregnancy in premature ovarian failure
The authors would like to acknowledge and appreciate Li
after therapy using Chinese herbal medicine. Chang Gung Medical Journal
XY for providing statistical insight. This work was supported by 2003;26:449–52.
research grants from the National Science & Technology Pillar [15] Higgins JP T. Green S. Cochrane Handbook for Systematic Reviews of
Program during the 12th Five-year Plan Period (No. 2012BAI Interventions Version 5.1.0. The Cochrane Collaboration, chapter 8; 2011.
32B04) and the Zhejiang Provincial Natural Science Foundation http://www.cochrane.org/training/cochranehandbook.
[16] Cao HN. Clinical study of the effect of er’zhi dihuang decoction about hor-
of China (No. LY13H040002).
mone levels and anti-ovarian antibodies in patients with premature ovarian
failure [in Chinese]. Journal of New Chinese Medicine 2009;41:53–4.
References [17] Ding Q, Wu ZY, Nie L. Clinical study of huluan decoction on premature
ovarian failure of heart–kidney imbalance [in Chinese]. Journal of TCM
[1] Committee opinion no. 502: primary ovarian insufficiency in the adoles- University of Hunan 2010;30:184–6.
cent. Obstetrics and Gynecology 2011;118:741–5. [18] Gao H, Xia T, Han B, Yang J. Clinical study of bushen tiaochong recipe
[2] Coulam CB, Adamson SC, Annegers JF. Incidence of premature ovarian on treating premature ovarian failure [in Chinese]. Liaoning Zhong Yi Za
failure. Obstetrics and Gynecology 1986;67:604–6. Zhi 2007;34:1557–60.
[3] Goswami D, Conway GS. Premature ovarian failure. Human Reproduction [19] Han LY, Ye YZ, Zhu PQ, Xing HM, Yan JY, Chu JZ, et al. Clinical study of
Update 2005;11:391–410. treating 36 cases of premature ovarian failure with kangluanshuai granules
[4] Zhou K, Jiang J, Wu J, Liu Z. Electroacupuncture modulates reproductive [in Chinese]. Liaoning Zhong Yi Za Zhi 2008;35:404–5.
hormone levels in patients with primary ovarian insufficiency: results from [20] Liang Z, Wang X, liu YX. Clinical study of kangshuai fuchao decoction
a prospective observational study. Evidence-Based Complementary and on treating premature ovarian failure [in Chinese]. Liaoning Zhong Yi Za
Alternative Medicine: eCAM 2013;2013:657234. Zhi 2010;37:1978–9.
[5] Nelson LM. Clinical practice. Primary ovarian insufficiency. The New [21] Ning Y, Ling QX, Shan ZQ, Ning Y, Huang JM. Clinical study of bushenyi-
England Journal of Medicine 2009;360:606–14. chong pill on premature ovarian failure of kidney deficiency [in Chinese].
[6] Panay N, Kalu E. Management of premature ovarian failure. Best Practice Hebei Medicine 2010;16:129–32.
& Research Clinical Obstetrics & Gynaecology 2009;23:129–40. [22] Yang XT. Clinical study of bushen decoction in treating premature ovar-
[7] Woad KJ, Watkins WJ, Prendergast D, Shelling AN. The genetic basis of ian failure [in Chinese]. Chinese Journal of Clinical Research 2011;24:
premature ovarian failure. Australian and New Zealand Journal of Obstet- 157–8.
rics and Gynaecology 2006;46:242–4. [23] Zhao RX. Clinical study of jiawei er’xian decoction in treating prema-
[8] Cust MP. Changes in practice since the publication in 2007 of further ture ovarian failure [in Chinese]. Chinese Journal of Information on TCM
results from the Women’s Health Initiative and the Million Women Study. 2010;17:62–3.
Menopause International 2007;13:141–3. [24] Zhu ZZ, Xiao PM. The efficiency of heche zibu decoction in treating pre-
[9] Zhu L, Luo SP, Xu LM, Zhu CL. Influence of Zuogui pill on antiovar- mature ovarian failure and its effect about hormone levels [in Chinese].
ian antibodies in immune premature ovarian failure rats [in Chinese]. Strait Pharmaceutical Journal 2011;23:149–50.
Modern Journal of Integrated Traditional Chinese and Western Medicine [25] Chattopadhyay D, Sen MR, Katiyar P, Pandey LK. Antiovarian anti-
2006;15:435–8. body in premature ovarian failure. Indian Journal of Medical Sciences
[10] Zhu L, Luo SP, Xu LM, Xie YH, Ren MN. Effects of Zuogui pills on 1999;53:254–8.
expression of Bcl-2 and Bax protein mice ovaries with immune premature

Please cite this article in press as: Wu Y, et al. Chinese Herbal Medicine for premature ovarian failure: A systematic review and meta-analysis.
Eur J Integr Med (2014), http://dx.doi.org/10.1016/j.eujim.2014.01.008

You might also like