Herbal medicine for the management of polycystic ovary syndrome (PCOS) and associated oligo/amenorrhoea and hyperandrogenism; a review of the laboratory evidence for effects with corroborative clinical findings

Background Polycystic ovary syndrome (PCOS) is a prevalent, complex endocrine disorder characterised by polycystic ovaries, chronic anovulation and hyperandrogenism leading to symptoms of irregular menstrual cycles, hirsutism, acne and infertility. Evidence based medical management emphasises a multidisciplinary approach for PCOS, as conventional pharmaceutical treatment addresses single symptoms, may be contra-indicated, is often associated with side effects and not effective in some cases. In addition women with PCOS have expressed a strong desire for alternative treatments. This review examines the reproductive endocrine effects in PCOS for an alternative treatment, herbal medicine. The aim of this review was to identify consistent evidence from both pre-clinical and clinical research, to add to the evidence base for herbal medicine in PCOS (and associated oligo/amenorrhoea and hyperandrogenism) and to inform herbal selection in the provision clinical care for these common conditions. Methods We undertook two searches of the scientific literature. The first search sought pre-clinical studies which explained the reproductive endocrine effects of whole herbal extracts in oligo/amenorrhoea, hyperandrogenism and PCOS. Herbal medicines from the first search informed key words for the second search. The second search sought clinical studies, which corroborated laboratory findings. Subjects included women with PCOS, menstrual irregularities and hyperandrogenism. Results A total of 33 studies were included in this review. Eighteen pre-clinical studies reported mechanisms of effect and fifteen clinical studies corroborated pre-clinical findings, including eight randomised controlled trials, and 762 women with menstrual irregularities, hyperandrogenism and/or PCOS. Interventions included herbal extracts of Vitex agnus-castus, Cimicifuga racemosa, Tribulus terrestris, Glycyrrhiza spp., Paeonia lactiflora and Cinnamomum cassia. Endocrine outcomes included reduced luteinising hormone (LH), prolactin, fasting insulin and testosterone. There was evidence for the regulation of ovulation, improved metabolic hormone profile and improved fertility outcomes in PCOS. There was evidence for an equivalent effect of two herbal medicines and the pharmaceutical agents bromocriptine (and Vitex agnus-castus) and clomiphene citrate (and Cimicifuga racemosa). There was less robust evidence for the complementary combination of spirinolactone and Glycyrrhiza spp. for hyperandrogenism. Conclusions Preclinical and clinical studies provide evidence that six herbal medicines may have beneficial effects for women with oligo/amenorrhea, hyperandrogenism and PCOS. However the quantity of pre-clinical data was limited, and the quality of clinical evidence was variable. Further pre-clinical studies are needed to explain the effects of herbal medicines not included in this review with current clinical evidence but an absence of pre-clinical data. Electronic supplementary material The online version of this article (doi:10.1186/1472-6882-14-511) contains supplementary material, which is available to authorized users.


Background
Polycystic ovary syndrome (PCOS) is a complex, common reproductive and endocrine disorder affecting up to 17.8% of reproductive aged women [1]. Medical management places strong emphasis on a multidisciplinary approach as pharmaceutical treatments appear to be only moderately effective in treating individual symptoms [2,3]. Conventional pharmaceutical management is limited by the prevalence of contraindications in women with PCOS [3], non-effectiveness in some circumstances [4], side effects [5] and by preferences of women with PCOS for alternatives to pharmaceutical management [6]. This review examines the mechanisms of effect for a potential alternative treatment, herbal medicine, and reveals six herbal medicines with both pre-clinical and clinical data explaining the reproductive endocrinological effects in PCOS and associated oligo/amenorrhoea and hyperandrogenism.
PCOS is a life-long condition and although the exact cause is yet to be identified, it is believed to have epigenetic origins, influenced by the uterine environment and behavioural factors [19]. Being overweight exacerbates all aspects of PCOS due to underlying metabolic disturbances [3]. Signs and symptoms are mediated by hormonal disorder including elevated androgens and fasting insulin, and abnormal relative ratio of the gonadotropins luteinising hormone (LH) and follicle stimulating hormone (FSH) [19]. Endocrine imbalances occur within the framework of disordered ovarian folliculogenesis, chronic anovulation, clinical signs of hyperandrogenism and metabolic syndrome [19].
Pharmaceutical treatment for menstrual irregularity includes the oral contraceptive pill (OCP) and ovulation induction with clomiphene citrate (clomiphene) [20,21] depending on fertility needs. Women with PCOS are however likely to exhibit contraindications for the OCP [3] and whilst induction of ovulation with clomiphene has demonstrated success, pregnancy rates remain inexplicably low [4]. Up to thirty 30% of women, particularly overweight women with PCOS, fail to respond to clomiphene therapy [4,22,23]. Management for hyperandrogenism includes anti-androgens and hypoglycaemic pharmaceuticals such as metformin [24]. Metfomin has demonstrated effectiveness for improving insulin sensitivity and hyperandrogenism, however use of metformin is associated with the high incidence of adverse effects including nausea, vomiting and gastro-intestinal disturbances [5].
Herbal medicines are complex interventions with the potential for synergistic and antagonistic interactions between compounds [25]. Effects within the body may also exhibit complexity by simultaneous interactions with various body systems, both biochemically and by altering organ function [26]. The focus of this review was studies investigating whole herbal medicine extracts with direct effects on reproductive endocrinology for the treatment of women with irregular menstruation, hyperandrogenism and PCOS. The rationale for using this methodology was to identify herbal medicines with current scientific evidence explaining specific reproductive endocrinological effects in PCOS, oligo/amenorrhoea and hyperandrogenism, to develop understanding for the direct effects of herbal medicines on reproductive endocrinology and to highlight herbal medicines for which there was current scientific evidence supporting herbal medicine selection. The purpose of this review is to inform clinical decisions in integrative settings and meet clinicians and consumers preferences for pragmatic herbal management within an holistic, individualised treatment frame [27,28].
We compared laboratory evidence including scientific studies using cell culture and animal models, with clinical data for proof-of-concept effects. A narrative synthesis of pre-clinical studies explaining reproductive endocrinological effects for herbal medicines with corroborative clinical evidence is presented.

Methods
We used the following definitions. PCOS according to the Rotterdam diagnostic criteria, specified by the presence of two out of three features; oligo/amenorrhoea, hyperandrogenism and polycystic ovaries on ultrasound [29,30]. Associated endocrine features for PCOS included elevated LH [31], which is strongly associated with infertility (p = 0.0003) [32] and miscarriage [33] and elevated fasting glucose which is prevalent in approximately 31% of women with PCOS including normal weight women [34].
Oligomenorrhoea was defined as menstrual cycle length that extended beyond 35 days (day one being the first day of menses). Amenorrhoea was defined as no menstrual period for three to six months or more [19]. This review was focussed on hypothalamic, pituitary and ovarian causes of menstrual irregularity with associated elevated gonadotropins including LH and prolactin and arrested folliculogenesis typically observed in polycystic ovaries. Hyperprolactinaemia is usually considered a unique cause for oligo/amenorrhoea; however in the present case it was included due to the potential co-existence for elevated prolactin, LH and PCOS, [32,35].
Hyperandrogenism was defined as clinical or biochemically excessive androgens. Clinical markers in females include cutaneous manifestations such as the presence of acne, hirsutism and/or male pattern alopecia. Biochemical indications include elevated plasma concentration of androgens.
We conducted two searches. The first was sensitive and aimed to capture all pre-clinical studies explaining the reproductive endocrine effects of whole herbal extracts in PCOS or associated oligo/amenorrhoea and hyperandrogenism. The second search was specific and sought only clinical studies investigating herbal medicines revealed during the first search (for which a mechanism of effect had been demonstrated). We additionally searched, on a case by case basis for pre-clinical evidence for herbal medicines identified during the second search, but not included in the results of the first search. Clinical studies were excluded based on the absence of evidence for a mechanism of effect for the whole herbal extract in reproductive endocrinology associated with PCOS, oligo/amenorrhoea and hyperandrogenism. We used this approach to improve transparency and to limit confirmation bias for herbal medicines favoured by the authors in clinical practice.
The first search revealed ten herbal medicines with a demonstrated mechanism of reproductive endocrinological effect for the whole herbal extract in PCOS, oligo/amenorrhoea and hyperandrogenism. These were Cimicifuga racemosa, Cinnamomum cassia, Curcuma longa, Glycyrrhiza spp., Matricaria chamomilla, Mentha piperita, Paeonia lactiflora, Silybum marianum, Tribulus terrestris and Vitex agnus-castus. Herbal medicines with a demonstrated mechanism of effect were entered as key terms in the second search.
We searched the following electronic databases: the Cochrane Library, MEDLINE ovidSP, CINAHL (1936 to present), SciVerse, EMBASE, PubMed, from the date of database inception to June 2014. In addition, we manually searched bibliographies of review articles.
Our laboratory search included investigations into the effects of herbal medicine using computer models, cell cultures, animals with PCOS induced with oestradiol valerate and androgens and sterilised and ovariectomised rats. We excluded laboratory studies which commenced using isolated chemicals not directly extracted from crude herbal medicines and studies examining androgen effects in male animals.
Our second search for clinical trials was performed without language restriction and included randomised controlled trials, non-randomised, open label and single arm clinical trials. We included clinical studies investigating commercially available herbal extracts and investigations that compared the effectiveness of herbal medicine with pharmaceuticals. We excluded clinical studies investigating herbal medicines with unrelated outcomes (including premenstrual syndrome, endometriosis and mastalgia) and clinical studies examining the effectiveness of complex herbal formulas for PCOS and associated oligo/amenorrhoea and hyperandrogenism, without demonstration of a mechanism of effect for the whole complex formula. We compared data from laboratory and animal studies with the outcomes of clinical trials. Clinical studies were assessed for risks of bias at study and outcome levels with risks summarised, tabulated (Tables 1 and 2) and presented in contextual narrative.

Equivalence for ovulation induction
for Tribulus Terrestris and Clomiphene for women with oligo/anovular infertility [60].

Glycyrrhiza glabra (European liquorice)
Two preclinical studies investigated the effects of Glycyrrhiza spp. for steroid hormone concentration and in polycystic ovaries. 2. Reduced free and total testosterone [53].
5. Improved ovulation rates in polycystic ovaries [50]. One laboratory study examined the effects for the combination Paeonia lactiflora and Glycyrrhiza uralensis on testosterone, oestradiol, FSH and LH in sterilised female rats [53].
Two single arm clinical trials examined androgen concentrations Following treatment with Paeonia lactiflora and Glycyrrhiza uralensis in the Chinese herbal combination Shakuyaku-kanzo-to. One included infertile oligomenorrhoeic women with hyperandrogenism (n = 8) [58] and the other included women with oligo/amenorrhoea and PCOS (n = 34) [59].

Paeonia lactiflora in combination with
Cinnamomum cassia Aqueous extract Unkei-to Paeonia lactiflora and Cinnamomum cassia combination was investigated for steroid hormonal effects on cultured human granulosa cells (obtained from women undergoing IVF). Cells were incubated with different doses for 48 hours [42] One clinical trial investigated the effects of Paeonia lactiflora and Cinnamomum cassia combination (Unkei-to) [57]. This single arm study included amenorrheic women aged 17-29 years (n = 157) with a sub group of women with hyper-functioning oligo/amenorrhoea (n = 42). Ovulation occurred in 61.3% of primary amenorrheic women and in 27.3% of secondary amenorrheic women following two months of treatment [57].
No detail for diagnostic criteria for PCOS. Confounding fertility factors not described. Drop-out reasons were not reported seven in Cimicifuga racemosa group and four in clomiphene group.
Women with PCOS and infertility, n = 194. Randomisation for 206 women 12 were excluded due to failure to respond (treatment group n = 7, control n = 5).
The miscarriage rate per pregnancy for the clomiphene alone group was 5 out of 33 (15.2%) and 6 out of 71 (8.5%) in the clomiphene plus Cimicifuga racemosa group.  isolated herbal chemicals (n = 3); inclusion of male subjects (n = 4); no pre-clinical evidence (n = 11) and conditions different to those specified (n = 13).

Excluded studies
Details of excluded studies investigating herbal medicines with clinical evidence but no preclinical evidence were provided in Table 3. Herbal medicines with preclinical evidence but no clinical evidence were provided in Table 4, and investigations into isolated chemicals derived from herbal medicine were presented in Table 5. Seven RCTs examined commercially produced herbal medicine extracts. These were Vitex agnuscastus in the form of Strontan® [64], Mastodynon® [61], Phyto Hypophyson® [62] and Agnacaston® [63] and Cimicifuga racemosa in the form of Klimadynon® [65,67,68] (Table 2).

Herbal medicines with effects in oligo/amenorrhoea, hyperandrogenism and PCOS
The results of preclinical studies and clinical studies have been summarised together for each of the six herbal medicines.

Vitex agnus-castus
Pre-clinical and clinical evidence was found for Vitex agnus-castus for lowered prolactin, improved menstrual regularity and treatment of infertility. Vitex agnus-castus contains a variety of compounds which bind to dopamine type 2 (DA-2) receptors in the brain; reduce cyclic adenosine mono phosphate (cAMP) and lowered prolactin secretion (Table 1). This was demonstrated in studies using recombinant DA-2 receptor proteins, and basal and stimulated rat pituitary cell cultures [38][39][40][41]. Prolactin lowering effects were found in normal and ovariectomised rats [49]. Additional agonistic opiate effects were

Studies included in qualitative synthesis Pre-clinical studies n = 18
Clinical studies n=15 Pre-clinical studies excluded (total7) for the following reasons;investigation of effects in male animals (n=4) and study commenced with isolated herbal chemicals (n=3). Clinical studies excluded (total 7) for the following reasons; investigated isolated herbal chemicals (n = 3), including male subjects (n=4). observed in studies using human opiate receptors cell cultures [70]. Clinical equivalence for prolactin lowering effects of Vitex agnus-castus (Agnucaston® 40 mg per day) and the pharmaceutical Bromocriptine (Parlodel® 5 mg per day) was found in one study including 40 women with hyperprolactinaemia [63]. Mean concentrations for prolactin following three months treatment with Vitex agnuscastus was significantly reduced from 946 mIU/l (±173) to 529 mIU/l (±297) (p < 0.001). Comparatively, mean prolactin concentration in the Bromocriptine group was significantly reduced from 885 mIU/l (±178) to 473 mIU/l (±266) (p < 0.001) demonstrating that both treatments were effective treatment for women with hyperprolactinaemia (normal reference range 25-628 mIU/l). The mean difference in prolactin reduction of the two groups was not significant (p = 0.96) ( Table 2).
Positive effects for Vitex agnus-castus in oligo/amenorrhoea and infertility was demonstrated in three placebo controlled RCTs [61,62,64]. In a study including Hormone concentration in obese women with PCOS [71]. Isolated constituent (epigallocatechin gallate 1) examined [72]. No evidence found for effects for whole herbal extract in PCOS, oligo/amenorrhoea and hyperandrogenism.
No evidence for mechanism of effect found for PCOS, oligo/ amenorrhoea or hyperandrogenism.
No evidence for mechanism of effect in PCOS, oligo/ amenorrhoea or hyperandrogenism found.
No evidence for mechanism of effect in PCOS, oligo/ amenorrhoea or PCOS revealed.
No mechanism of effect in PCOS, oligo/amenorrhoea or hyperandrogenism found.
No evidence for mechanism of effect found in PCOS, oligo/ amenorrhoea or hyperandrogenism (in female cell cultures or animals).
No mechanism of effect in PCOS, oligo/amenorrhoea or hyperandrogenism.

Stachys lavandulifolia (wood betony)
Evidence for improved uterine bleeding (including oligomenorrhoea and amenorrhoea) in women with PCOS comparable with Medroxyprogesterone acetate [86].
No mechanism of effect studies found for whole herbal extract in PCOS and or associated oligo/amenorrhoea and hyperandrogenism.

Table 4 Herbal medicines with pre-clinical evidence not included in this review
Herbal medicine Pre-clinical evidence for (potential) effects in reproductive endocrinology in PCOS and associated oligo/amenorrhoea and hyperandrogenism
No clinical evidence examining effectiveness in women was found.
No clinical data found.
No clinical data for women.
No clinical evidence including women was found.
Studies investigating chemical compounds derived from the herbal medicines, included in this review but investigating different outcomes were found for Vitex agnus-castus [70] and Cimicifuga racemosa [106].
women with menstrual irregularity and infertility (n = 96), menstrual cyclicity was significantly improved for women treated with Vitex agnus-castus (Mastodynon® 30 drops per day for three months) compared to placebo (p = 0.023) [61] ( Table 2). Another study, including women with sub fertility (n = 67), showed improved menstrual cyclicity for a sub-group of women with oligomenorrhoea following treatment with Vitex agnuscastus (Phyto-Hypophyson® 7.5 ml per day) compared to placebo, (p = 0.023) [62] (Table 2) (Table 2). Methodological shortcomings included not reporting baseline characteristics for subgroups and small sample sizes; however clinical outcomes demonstrated physiological effects consistent with laboratory and animal findings (Tables 1 and 2).

Cimicifuga racemosa
Cimicifuga racemosa was found to lower LH in two laboratory studies both examining cell cultures from ovariectomised rats [45,52] (Table 1). The mechanism occurred through competitive inhibition of oestrogen following the selective binding of oestrogen receptors (ERα) on the hypothalamus and pituitary [52]. An earlier study found contrary results for reduction of LH, however this study investigated an isolated flavonoid and suggested that other constituents may be active [37]. Three RCTs corroborate the positive fertility effects for Cimicifuga racemosa in women with PCOS, used in conjunction and when compared with the pharmaceutical Clomiphene citrate (clomiphene), [65,68,71] (Table 2). Results were reported for 441 women and show improved pregnancy rates when Cimicifuga racemosa was added to clomiphene during one menstrual cycle. In a study including women with PCOS (n = 147), pregnancy rates for the group receiving combined therapy (clomiphene 150 mg plus Cimicifuga racemosa 20 mg per day (Klimadynon®)) were 43.3% compared to 20.3% for women receiving only clomiphene [65] (Table 2). In another study using similar methodology (n = 100) pregnancy rates were 34.8% for the group treated with Cimicifuga racemosa plus clomiphene compared to 17.2% for women treated with clomiphene alone [68] (Table 2). Another study included women with PCOS and infertility (n = 100) compared Cimicifuga racemosa (Klimadynon®) and clomiphene over three months for hormone concentrations and pregnancy rates. Pregnancy rates were higher in the women in taking Cimicifuga racemosa compared to clomiphene, 14% and 8% respectively; however differences were not statistically significant. This study found significant effects for lowered luteinising hormone for women with PCOS receiving Cimicifuga racemosa compared to clomiphene (p = 0.007) [67]. Findings from clinical studies concur with laboratory and animal studies; however potential risks for bias include performance and collection bias due to lack of blinding ( Table 2).

Tribulus terrestris
Two laboratory based RCT's examined the effects of Tribulus Terrestris in rats with polycystic ovaries induced with oestradiol valerate [46,47] (Table 1). Both studies demonstrate significantly improved ovulation rates for animals treated with two doses of Tribulus terrestris extracts compared to controls. Although the endocrinological effects were not described in either study, laboratory findings of ovulation induction are supported by the clinical findings of elevated FSH following treatment with Tribulus terrestris [56] (Table 2).

Isolated chemicals
Evidence for effects
with caution due to risks for bias in clinical studies. One study was uncontrolled with a small number of healthy participants [56], the second study did not report baseline characteristics, methods for allocation to treatment groups and data were not statistically analysed [60] ( Table 1).

Glycyrrhiza spp
Androgen lowering effects for Glycyrrhiza spp. have been demonstrated in one laboratory study examining hormone concentration in female rats (Glycyrrhiza uralensis), [53] and corroborated in two clinical trials, one including healthy women [55] and the other including women with PCOS (Glycyrrhiza glabra) [54] ( Table 1). The animal study reported significantly reduced free and total testosterone and increased oestradiol in sterilised rats and no hormonal changes in oophrectomised rats. The authors conclude that the hormonal effects occurred primarily in the ovary via enhanced aromatisation of testosterone to 17-beta oestradiol. The investigators also observed significantly increased oestradiol. There were no changes to FSH or LH in androgen sterilised or oophrectomised rats [53]. Another animal study examined the effects of Glycyrrhiza uralensis on the morphological features of polycystic ovaries using immunohistochemistry [50] (Table 1). This study demonstrated significantly increased ovulation rates by the number of corpus luteum in polycystic ovaries compared with controls. The authors propose that the mechanism of effect for Glycyrrhiza uralensis was competitive inhibition of oestrogen at oestrogen receptor sites, limiting the production of nerve growth factor (NGF), its neurotropic effects and inhibition of sympathetic neurological involvement in the pathogenesis of polycystic ovaries.
Two clinical studies examined the androgen lowering effects of Glycyrrhiza Glabra. A single arm clinical trial demonstrated reduced testosterone in healthy women aged 22-26 years (n = 9) over two menstrual cycles. Treatment with Glycyrrhiza glabra, 7 grams per day reduced testosterone from 27.8(±8.2) to 17.5 (±6.4), p < 0.05 [55]. Another single arm clinical trial investigated the effects of Glycyrrhiza glabra in women with PCOS, (n = 32). Glycyrrhiza glabra 3.5 g per day was added to anti-androgen pharmaceutical treatment, Spirinolactone 100 mg/day over two menstrual cycles. An unwanted side effect for Spirinolactone was the flare of androgens during the initial phase of treatment. This study demonstrated reduced concentrations of testosterone during the first four days of treatment at 103 ± 29 ng/d in the Spirinolactone group compared to 91 ng/d (±19) when combined with Glycyrrhiza glabra (p < 0.05) [54] (Table 1)

Paeonia lactiflora and glycyrrhiza uralensis
One laboratory study and two clinical investigations provided evidence for the two herb combination, Glycyrrhiza uralensis and Paeonia lactiflora [53,58,59] (Table 1). An animal study found significant reductions in free and total testosterone following exposure to the combination [53] ( Table 1). These findings were supported in two open label clinical trials including women with PCOS (n = 34) [59] and women with hyperandrogenism (n = 8) [58]. Both trials examined the effects on androgens for the aqueous extract TJ-68 (equal parts Glycyrrhiza uralensis and Paeonia lactiflora), 75 grams per day for 24 weeks and 5-10 grams per day for 2-8 weeks respectively. In the trial including women with PCOS, mean serum testosterone was significantly reduced from 137.1 ng/dL (±27.6) to 85.3 ng/dL (±38), p < 0.001 at four weeks of treatment [59]. Similar effects were observed in the women with oligomenorrhoea and hyperandrogenism which showed serum testosterone reduced from 50-160 ng/dL prior to treatment to less than 50 ng/dL [58]. However statistical significance was not reached due to the small sample size despite positive outcomes in seven out of eight participants (Table 1).

Paeonia lactiflora and cinnamomum cassia
Paeonia lactiflora combined with Cinnamomum cassia in a preparation called Unkei-to was investigated in an invitro study for ovarian production of 17-beta-oestradiol and progesterone, [42] (Table 1). Granulosa cells obtained from women undergoing IVF were examined for steroid hormone concentration following incubation with different doses over 48 hours. Oestradiol was significantly increased (p < 0.01) following exposure to doses of 0.3 ug/ml of Unkei-to. Supporting clinical evidence was found in one clinical trial of 157 infertile women aged 17-29 years, including a subgroup of 42 women with hyperfunctioning (PCOS) oligo/amenorrhoea. Treatment with Unkei-to, 7.5 grams per day for eight weeks, demonstrated significant reductions of mean LH in the PCOS sub-group of 49.7% (±15.3). Ovulation was confirmed in 30 out of 42 oligo/amenorrheic women [57] (Table 1). Limitations however include findings based on sub-group comparisons without description of subgroup baseline characteristics (other than oligomenorrhoea). Although the same aqueous extract intervention was investigated in preclinical and clinical studies, it contained additional herbal extracts and it was irrational to attribute hormonal effects to Paeonia lactiflora and Cinnamomum cassia.

Cinnamomum cassia
An animal study compared the effectiveness of Cinnamomum cassia and the pharmaceutical Metformin on hormone concentration in rats with PCOS [48] (Table 1). Both interventions demonstrated significant improvements compared to controls at 15 days for measures of testosterone ng/ml (control 0.747 ± 0.039; metformin 0.647 ± 0.027; Cinnamomum cassia 0.625 ± 0.029); LH ng/ml (control 7.641 ± 0.267; metformin 6.873 ± 0.214; Cinnamomum cassia 6.891 ± 0.221) and insulin resistance (HOMA-IR) (control 10.018 ± 0.217; metformin 7.067 ± 0.184 Cinnamomum cassia 8.772 ± 0.196) (p < 0.05) [48]. The metabolic effects for Cinnamomum cassia were further demonstrated in overweight women with oligo/amenorrhoea and PCOS in a placebo controlled RCT [66] (Table 2). However, although the RCT had low risks for bias, it was a pilot study primarily investigating feasibility. Outcomes were promising for metabolic profile in PCOS however the sample size was small and the authors recommended further studies.

Summary of results
This review includes 18 preclinical laboratory based studies and 15 clinical trials. We found reproductive endocrine effects in oligo/amenorrhoea, hyperandrogenism and/or PCOS for six herbal medicines. The quality of evidence, as determined by the volume of pre-clinical studies and the methodological quality of clinical trials, was highest for the herbal medicines Vitex agnus-castus, Cimicifuga racemosa and Cinnamomum cassia, for which there were laboratory and/or animal studies demonstrating endocrine mechanisms of action consistent with clinical outcomes shown in RCT's with low risks for bias. However, replicated RCT data was only found for one herbal medicine, Cimicifuga racemosa.
Evidence for Tribulus terrestris, Glycyrrhiza spp. alone and in combination with Paeonia lactiflora and Paeonia lactiflora with Cinnamomum cassia was limited by the volume of laboratory and animal studies, with only one to two studies found for each herb or herbal combination. There was supporting clinical data, however many were small single arm, open label studies measuring endocrine effects in healthy women. Evidence for these herbal medicines is preliminary and in an emergent phase.

Discussion
This review synthesises the evidence for mechanisms of effect for herbal medicine in oligo/amenorrhoea, hyperandrogenism and PCOS. Laboratory, animal and clinical studies demonstrate that the herbal medicines Vitex agnus-castus, Cimicifuga racemosa and Tribulus terrestris initiate endocrine effects in the pituitary as measured by lowered prolactin and LH and raised FSH. Four herbal medicines, Tribulus terrestris, Glycyrrhiza spp., (alone and in combination with Paeonia lactiflora), Paeonia lactiflora (in combination with Cinnamomum cassia) and Cinnamomum cassia demonstrated morphological changes in polycystic ovaries and steroidogenesis, including reduced ovarian volume and cysts, lowered androgens, improved insulin sensitivity and increased oestradiol.
Clinical investigations found no adverse effects for the six herbal medicines included in this review ( Table 2). A comparative study investigating the pharmaceutical Bromocriptine and the herbal medicine Vitex agnuscastus found no side effects associated Vitex agnus-castus compared to 12.5% of participants taking Bromocriptine reporting nausea and vomiting [63]. No studies comparing the effectiveness for herbal medicines and the oral contraceptive pill in PCOS, oligo/amenorrhoea and hyperandrogenism were found.
Herbal medicine may present a treatment option for women with oligo/amenorrhoea, hyperandrogenism and PCOS as an adjunct or alternative treatment to pharmaceuticals with a high degree of acceptability by women with PCOS [6]. Preliminary evidence for equivalent treatment effects were found for the two pharmaceuticals and three herbal medicines. These were bromocriptine, in the management of hyperprolactinaemia andVitex agnuscastus and clomiphene for infertility and ovulation induction and Cimicifuga racemosa and Tribulus terrestris. Herbal medicine had positive adjunct effects with the pharmaceuticals Spirinolactone in the management of hyperandrogenism (Glycyrrhiza Spp.), and clomiphene for PCOS related infertility (Cimicifuga racemosa). It is important however to highlight that evidence was provided by a limited number of clinical studies, some with significant risks for bias; particularly Tribulus terrestris, Glycyrrhiza glabra alone and in combination with Paeonia lactiflora and Paeonia lactiflora in combination with Cinnamomum cassia.
Selection of herbal medicines for the management of PCOS often includes the combined prescription of Glycyrrhiza spp. and Paeonia lactiflora [72][73][74][75]. We found preliminary evidence for this combination for hyperandrogenism only, and the evidence was more robust for Glycyrrhiza spp. alone than when combined with Paeonia lactiflora. Comparatively, our findings for the combination of Peaonia lactiflora and Cinnamomum cassia demonstrated no change in androgen concentration, suggesting that the anti-androgen activity in the Glycyrrhiza spp. and Paeonia lactiflora combination more likely attributable to Glycyrrhiza spp. However our findings may be complicated by the aqueous extraction methods used in the Paeonia lactiflora and Cinnamomum cassia combination and the preclinical studies into the Glycorrhizza spp and Paeonia lactiflora combination. More research into the antiandrogen effects of the combination Glycyrrhiza spp. and Paeonia lactiflora is needed to clarify the antiandrogen mechanism particularly if this herbal combination remains cornerstone herbal management for hyperandrogenism.
This review has some limitations. We used a methodological approach which was deductive and not consistent with traditional rationale for herbal selection. Our inclusion criteria for clinical studies were specific and relied upon our identification of herbal medicines with preclinical (laboratory based) evidence explaining the mechanisms of reproductive endocrinological effects in oligo/amenorrhoea, hyperandrogenism and PCOS. Clinical studies were excluded from this review due to the absence of evidence for whole herbal extracts. This was the case for Camellia sinensis (green tea) for which only one laboratory study investigated the effects of injecting epigallocatechin, a catechin found in green tea in animals [76]. High quality clinical evidence for Camellia sinensis was not presented in this review due to the absence of pre-clinical data explaining the mechanism for effect for the whole herbal extract [77]. Mentha spicata (spearmint) was another herbal medicine excluded from this review despite the availability of high quality clinical evidence demonstrating testosterone lowering effects in women with PCOS [78]. We found no laboratory evidence describing the mechanism of action for Mentha spicata in hyperandrogenism. Camilla sinensis and Mentha spicata are examples of herbal medicines excluded from this review due to not meeting the inclusion criteria. Studies investigating western herbal medicines excluded from this review are provided in Tables 3, 4

and 5.
Our search strategy may have restricted access due to limited search terms. We didn't include alternative spelling of oestrogen and additional search terms for herbal medicine could have been included to increase sensitivity of the search.
This study synthesises the evidence for reproductive endocrine effects for six whole herbal medicine extracts that may be used to treat PCOS and associated oligo/ amenorrhoea and hyperandrogenism. The findings were intended to add to clinicians understanding for the mechanisms of action for herbal medicine for treatment in these common conditions and reveal herbal medicines with reproductive endocrinological effects, currently demonstrated in scientific literature.

Conclusions
Preclinical and clinical studies provide preliminary evidence that six herbal medicines may have beneficial effects for women with oligo/amenorrhea, hyperandrogenism and PCOS. The quality of the evidence is variable and strongest for Vitex agnus-castus and Cimicifuga racemosa in the management of oligo/amenorrhea and infertility associated with PCOS; and Cinnamomum cassia for improving metabolic hormones in PCOS. Evidence for Tribulus terrestris, Glycyrrhiza spp. alone and in combination with Paeonia lactiflora and Paeonia lactiflora combined with Cinnamon cassia is promising but in an emergent phase. Further investigations into the mechanisms of effect for herbal extracts are needed to complete our understanding of the reproductive endocrinological effects for herbal medicine for these common conditions.