Re: Meta-analysis Finds Chaste Tree Effective for Premenstrual Syndrome, but Published Research Is Heterogeneous; More High-quality Studies Are Needed

Chaste Tree (Vitex agnus-castus; Lamiaceae) Premenstrual Syndrome Systematic Review/Meta-analysis Date: 06-30-2018 HC# 111763-595 Verkaik S, Kamperman AM, van Westrhenen R, Schulte PFJ. The treatment of premenstrual syndrome with preparations of Vitex agnus-castus: a systematic review and meta-analysis. Am J Obstet Gynecol. August 2017;217(2):150-166. doi: 10.1016/j.ajog.2017.02.028. Premenstrual syndrome (PMS) is the occurrence of physical and psychological symptoms during the luteal phase of the menstrual cycle and continuing into the first few days of menstruation. Symptoms include headache, edema, breast or abdominal pains, anxiety, depression, agitation, and aggression. Mild symptoms of these types are normal for many women, but in 5-8% of women, they can be severe and may be characterized as premenstrual dysphoric disorder (PMDD). While there are many hypotheses as to the cause of PMS and PMDD, none have been scientifically proven. The most common pharmaceutical treatment prescribed is selective serotonin reuptake inhibitors (SSRIs). These can have many undesired side effects such as anxiety, weight gain, nausea, sexual dysfunction, and more. Chaste tree (VAC; Vitex agnus-castus; Lamiaceae) fruit is among the alternative remedies commonly used to treat PMS symptoms. The goal of this systematic review and meta-analysis was to determine the efficacy and tolerability of chaste tree in the treatment of PMS and/or PMDD. Ten databases were searched including MEDLINE, Embase, Web of Science, PubMed, and Google Scholar, using search terms such as "premenstrual syndrome," "PMS," and variations, as well as "vitex," "chaste tree," and other common names in multiple languages. The search was updated in January of 2016. Clinical studies of women of reproductive age diagnosed with PMS or PMDD were considered. Included studies were randomized controlled trials that used any dose or preparation of chaste tree, with either a placebo or pharmacotherapy comparator and a minimal duration of two menstrual cycles. Studies using homeopathic preparations of chaste tree or combination treatments were excluded. Outcomes data preferred for meta-analysis were those based on validated scales. Discontinuation rates, adverse events, and satisfaction were also analyzed. Risk of bias was assessed for each trial using the Cochrane Risk of Bias Tool, and potential publication bias was determined using a funnel plot and Egger test. After title and abstract screening of 395 articles, 38 articles were selected and the full-text was assessed for eligibility. Seventeen randomized controlled trials were ultimately included in the qualitative analysis, 14 of which were also included in the quantitative analysis; sufficient data were not available for the other three. Eight articles were written in English, four in Farsi, three in Italian, and two in Turkish. The authors report using a translator as necessary to interpret these. Ten articles were placebo controlled and nine were comparator controlled, with comparators including the SSRI, fluoxetine, an oral contraceptive, pyridoxine, magnesium, and St. John's wort (Hypericum perforatum, Hypericaceae) or vitamin E. Two trials compared chaste tree to both an oral contraceptive and a placebo. Most studies were double-blind, though two were single-blind and blinding was not adequately explained in two others. Seven preparations of VAC were specified in 11 studies. Six preparations were extracts and one was capsules containing ground dried berries. Six studies did not specify the preparation. In 10 of the studies that used a placebo control, , VAC was found to be superior in all but one.. The one negative study used ground berries, where all others used extracts. VAC was found to be superior for relieving PMS symptoms in studies comparing it to pyridoxine, magnesium, St. John's wort, or vitamin E. VAC was compared to oral contraceptives in three studies and was found to be comparable. VAC was compared to fluoxetine in two studies. In one study, VAC was comparable to fluoxetine. In the other, fluoxetine was superior for some components of the Hamilton depression rating scale. The authors rated only two studies as having a low risk of bias, six were considered to have a moderate risk of bias and nine a high risk, with reporting incomplete for all studies. The pooled effect of VAC vs. placebo for all PMS symptom was large (Hedges g -1.21; Hedges g; 95% confidence interval [CI], -1.53 to -0.88; I2, 91%). Meta-analyses of studies separately reporting depressive symptoms (n=5) and anxiety symptoms (n=2) also showed a large effect. There was no significant difference between the efficacy of VAC and either oral contraceptives or fluoxetine, but fewer adverse events occurred with VAC than with either of these comparators. Adverse events in placebo-controlled studies were mild and not substantially different from placebo, and the products were considered well-tolerated. Dropout rates varied greatly among studies but did not differ for VAS vs. placebo. Heterogeneity among studies was very high, and both the funnel plot and Egger tests suggested the possibility of publication bias. Smaller studies reported larger effect sizes, leading to the suspicion that small studies with negative results might exist that have not been published. Among analyses of heterogeneity, it should be noted that double-blind studies reported greater effects than single-blinded or poorly described studies (P<0.05); although, studies suspected of reporting bias also reported greater effects (P value not stated), and those using commercial products reported greater effects than those using custom-made products (P=0.036). The authors conclude that while the meta-analysis shows a large pooled effect for VAC in placebo-controlled trials, they are not convinced a definitive conclusion can be drawn on its effectiveness due to the high risk of bias, high heterogeneity, and risk of publication bias. There is a clear need for high-quality research of appropriate duration and adequate number of subjects. The authors suggest these be conducted using standardized extracts, and could include both superiority trials vs. placebo and non-inferiority trials with SSRIs and oral contraceptives as comparators. The first author works for a biotechnology corporation that does not presently work on drugs for PMS/PMDD and was not involved in the study. The other authors report no potential conflicts. —Erin Smith, MSc., CCH