A Systematic Review and Meta-Analysis of Acarbose in the Treatment of Polycystic Ovary Syndrome

Research Article

J Endocr Disord. 2014;1(3): 1013.

A Systematic Review and Meta-Analysis of Acarbose in the Treatment of Polycystic Ovary Syndrome

Lihua Wang, Tingting Han, Jiang Yue, Wei Liu and Yaomin Hu*

Department of Endocrinology and Metabolism diseases, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, China

*Corresponding author: Yaomin Hu, Department of Endocrine and Metabolic diseases, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, Shanghai 200127, China

Received: August 18, 2014; Accepted: October 08, 2014; Published: October 10, 2014


Objective: To evaluate the efficacy and tolerability of acarbose in the treatment of Polycystic Ovary Syndrome (PCOS).

Methods: We researched the database of PubMed, Biomedical Core Database (BCD), Cochrane library databases, and Chinese Biological Medicine for Randomly Control Trials (RCTs) and analyzed the retrospect data of acarbose in the treatment of PCOS.

Results: There are seven RCTs that met the inclusion criteria. As different control groups (placebo or metformin) have been used in these trials, we separated the data into two independent groups. The dose of acarbose in the treatment groups ranged from 150 to 300 mg/day. VLDL, triglycerides, Luteinizing Hormone (LH), testosterone, and Dehydroepiandrosterone Sulfate (DHEAS) levels decreased, whereas HDL significantly increased in acarbose groups when compared with the placebo groups. Acarbose treatment also improved menstrual frequency and insulin level. We couldn’t find any differences between acarbose group and metformin group in fasting insulin level, BMI, LH, FSH, and testosterone level. However, based on these three RCTs data, the pregnancy rate was higher in acarbose group than that in metformin group (OR=3.02, 95% CI: 1.10-7.71, P=0.02). The gastrointestinal adverse effects of acarbose were found to be higher in treatment groups than those in placebo groups, and similar to or slightly less than those in metformin groups.

Conclusions: Treatment with acarbose, at the dose of 150-300 mg/d, has been shown to improve various clinical manifestations of PCOS and be a safe and effective drug for the treatment of these patients, especially for patients who are intolerant of metformin. However, it may be too early to draw conclusions due to limitations in the currently available data.


Polycystic Ovary Syndrome (PCOS), which affects 2.2-10% of reproductive women, is a complex, multisystemic disorder presenting with oligomenorrhea of amenorrhea, infertility, hirsutism, acne, hypersecretion of androgens, and polycystic ovary [1-4]. PCOS is associated with obesity, insulin resistance, hyperinsulinemia, impaired glucose tolerance and disorders of lipid metabolism, suggesting that this subgroup of patients is at risk for the development of type 2diabetes and cardiovascular disease [5]. In addition, endometrium, ovary and breast cancers are also long-term consequences of PCOS [6,7]. Therefore, active invention for PCOS to prevent further problems is deemed to vital.

Insulin resistances, presented as impaired insulin-mediated glucose uptake, are an integral feature of PCOS and have been reported in both obese and non-obese patients [8,9]. It has been shown that hyperinsulinemia may enhance LH release and androgen synthesis and secretion, also inhibit hepatic synthesis of Sex Hormone Binding Globulin (SHBG). Up to 30% of obese PCOS patients showed impaired glucose tolerance and 7.5-10% patients were diagnosed of Type II Diabetes Mellitus (T2DM) screened by Oral Glucose Test (OGTT). Therefore, numbers of medications that improve insulin sensitivity have been used to treat PCOS patients.

Acarbose, and α-glucosidase inhibitor, is a common first-line treatment in T2DM. Acarbose is shown to reduce and slow down the intestinal absoption of glucose, which subsequently minimise the post-prandial rise of blood glucose and insulin concentration. Acarbose does not cause hypoglycemia and its minor gastrointestinal side effects can be prevented by gradual dosage increments. Therefore, acarbose has been used in the management of PCOS in recent years [10].

Recently, several clinical trials have been conducted to investigate the effects of acarbose on PCOS by comparing the effects with placebo or metformin. The objective of this meta-analysis is to pool data from these trials and trying to evaluate the efficacy and tolerability of acarbose in the treatment of PCOS.

Research Design and Methods

A comprehensive and systematic search of published literature for trials of acarbose in the treatment of PCOS was performed during May 2014 using PubMed, Biomedical Core Database (BCD), Cochrane Library Databases, and Chinese Biological Medicine. The search strategy was not limited by year or language of publication. The key words used in this search were acarbose or α-glucosidase inhibitor, and PCOS or polycystic ovary syndrome. Study treatment durations ranged from 3 to 6 months, and the effective dosing ranged from 150 to 300 mg/d. Each of the studies shared fundamental inclusion criteria, including: menstrual disorders (<6 menstruations/12 months), clinical [Ferriman-Gallwey (FG) index ≥8] or laboratory (testosterone >80 ng/dl and/or androstenedione >190 ng/dl) hyperandrogenism and insulin resistance. Alteration of hepatic, renal and thyroid function, presence of congenital adrenal hyperprolactinemia, presence of congenital adrenal hyperplasia, presence of diabetes, and the use of hormonal medications or medications that might interfere with carbohydrate metabolism over the last 6 months were exclusion criterions in this study. Six of seven trials studied patients with obesity.

The efficacy measures in each study were clinical (number of menstrual cycles, FG score, ovulation or pregnancy rate), anthropometric (weight, height, BMI), hormone and metabolic evaluation before and after treatment. Safety measures included incidence of adverse events, such as flatulence and/or diarrhea.

Statistical Analysis

A three-item, 1-5 quality scale to score each report has been used to meet the inclusion criteria. The use of concealment and intention-to-treat analysis was also assessed. Two of the three reviewers made quality assessment. Their disputes were settled by consensus. The analyzed data consisted of group means that were reported in one paper. The results were combined and expressed as Odds Ratio (OR) or Weight Mean Difference (WMD) with 95% confidence intervals (95% CIs) using a Fixed Effect (FE) or Randomized Effect (RE) model, for the studies with sufficient data. Additionally, homogeneity was assessed with I2 statistic and χ2 test. All statistics for meta-analysis were calculated by Revman Manager 5.2 Software (Copenhagen, Denmark). Sensitivity analysis was applied to explore the influence on outcomes via changing effect model or excluding studies with abnormal results.


Description of studies

Seventy citations have been screened for eligible for inclusion in meta-analysis. Sixty-three studies were excluded including duplicated data from different databases or not related (n=55), general reviews of drugs or disease areas (n=5), and self-control studies (n=2). Therefore, seven studies, which met the inclusion criteria, were used in our meta-analysis [11-17]. Six studies and an abstract of one study were published in English. As different drugs were used in control groups, we devided these studies into two groups. The characteristics and methodological quality of the included studies are shown in Table 1. The age of the subjects was not reported in some of the studies.