AIMS

To evaluate the effects of pioglitazone on menstruation and the metabolic parameters of non-obese women of reproductive age with polycystic ovary syndrome (PCOS) and compare the effects among different subgroups of PCOS patients with different clinical presentations.

METHODS

Twenty-eight women of reproductive age with PCOS were recruited; 20 women finished the study. The women were divided into three groups according to clinical presentations: group A (n = 4) had chronic oligo- or anovulation with polycystic ovaries; group B (n = 5) had chronic oligo- or anovulation with hyperandrogenism; and group C (n = 11) had chronic oligo- or anovulation, hyperandrogenism and polycystic ovaries. Pioglitazone (15 mg/day) was given for 6 months and the therapeutic effects were evaluated. Menstrual cycle regularity and hormone levels (plasma luteinizing hormone (LH), follicle-stimulating hormone, total cholesterol, triglycerides, high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol, glucose, insulin, C-peptide, free testosterone, homeostatic model assessment (HOMA)) were evaluated during and after pioglitazone treatment.

RESULTS

Thirty percent and 45.5% of the patients showed improvement of menstrual cycle regularity immediately after completion of treatment and at 6 months after completion of pioglitazone treatment, respectively, although there were no statistical differences among the subgroups. There was a significant difference in the change of body mass index (BMI) throughout the study period among the subgroups (p = 0.008). The decrease in BMI was significantly higher in group B than in groups A and C at 3 months (p = 0.0381) and 6 months of treatment (p = 0.0054), as well as at 6 months after completion of treatment (p = 0.003). HDL-C concentrations increased throughout the period (p = 0.001) without a difference among the subgroups. LH levels decreased at 6 months of treatment and throughout the follow-up period (p = 0.0045), but this did not differ among subgroups. The free testosterone level decreased, but without significance. There was no statistical improvement in any of the parameters of insulin resistance, but baseline free testosterone levels were related to the improvement of the HOMA insulin sensitivity score (p = 0.0009). Patients with more than a 50% decrease of their HOMA insulin resistance (HOMA-IR) score showed higher baseline free testosterone levels (hyperandrogenic groups B and C) than did the patients with less than 50% HOMA-IR score improvement or those patients with an increased HOMA-IR score.

CONCLUSIONS

These preliminary results suggest that pioglitazone treatment for non-obese PCOS women of reproductive age may be effective to help the resumption of the menstrual cycle, and the patients showed a different response pattern according to their baseline free testosterone levels.