Five to 8% of women of reproductive age suffer from moderate to severe symptoms of premenstrual syndrome associated with substantial distress or functional impairment.

To determine the efficacy, tolerability and acceptability of Vitex agnus castus (VAC, chaste-berry) preparations for treatment of premenstrual syndrome (PMS).

After an extensive literature search all randomized controlled trials of VAC in women with PMS and/or premenstrual dysphoric disorder were included.

We found 17 randomized controlled trials of VAC in the treatment of PMS of which 14 studies could be included in the quantitative analysis. All trials, except one, found the Vitex agnus castus preparation to be more efficacious than placebo, pyridoxine and magnesium for either total symptom score or individual symptoms or symptom clusters. Unfortunately, the majority of trials is hampered by a high risk of bias. The pooled effect of VAC in placebo-controlled trials was large (Hedge's g: −1.21; 95% CI: −1.53 to −0.88), but heterogeneity extremely high (I2 = 91%). We could not single out distinct factors that could explain this heterogeneity. The funnel plot and Egger's tests suggest the presence of bias over the full set of studies and effect sizes.

Although meta-analysis shows a large pooled effect of VAC in placebo-controlled trials the high risk of bias, high heterogeneity and risk of publication bias of the included studies preclude a definitive conclusion. The pooled treatment effects should be viewed as merely explorative and, at best, overestimating the real treatment effect of VAC for PMS symptomatology.

P.S. reports personal fees from H. Lundbeck A/S, outside the submitted work. The other authors have not supplied their declaration of competing interest.