Effects of thiazolidinediones on polycystic ovary syndrome: a meta-analysis of randomized placebo-controlled trials.

INTRODUCTION Thiazolidinedione (TZD) is one of the therapy options for polycystic ovary syndrome (PCOS) patients; however, the effectiveness of TZD in the treatment of PCOS remains controversial. The aim of this metaanalysis was to clarify the role that TZDs play in the treatment of PCOS. METHODS The authors searched the following databases for any date up to June 2012 for randomized controlled trials on PCOS treatment in which interventions for the experimental and control groups were TZDs and placebo, respectively: Medline, Embase, and the Cochrane library. RESULTS The search strategy identified 173 potential publications, eight of which were included. In the treatment of PCOS, the insulin-lowering effects of TZDs were superior to placebo (95% CI -1.50 to -0.12; P=0.021), and the lowering of fasting blood glucose was superior to placebo (95% CI -1.06 to -0.05; P=0.031). There was no difference in reduction of the Ferriman-Gallwey scores or the androgen levels between TZDs and placebo (95% CI -0.57 to 0.10; P=0.169 and 95% CI -0.64 to 0.09; P=0.141, respectively). The effects of TZDs on body weight reduction were inferior to placebo (95% CI 0.13 to 0.66; P=0.003). Significant between-study heterogeneity was detected for several variables assessed. CONCLUSION This is the first meta-analysis to evaluate the role that TZDs plays in the treatment of PCOS compared with placebo. The currently available data showed that TZDs can effectively reduce insulin and fasting blood glucose levels in patients with PCOS, but TZDs may not effectively reduce the Ferriman-Gallwey score or androgen levels and may increase body weight.