Metformin (Glucophage) has proved effective as an ovulatory agent in women with polycystic ovary syndrome (PCOS). This effect generally is attributed to a reduction of insulin resistance and hyperandrogenism, and the best response to metformin has been thought to occur in women with higher insulin or testosterone levels, or both. Two studies, however, show that metformin may induce ovulation in women with PCOS while effecting only minimal changes in insulin and androgen levels. A diagnosis of PCOS historically required hyperandrogenism, thus normoandrogenic women with anovulation were not included in studies assessing metformin’s effect on ovulation. In a randomized study, Carmina and Lobo evaluated the use of metformin for inducing ovulation in anovulatory women with no evidence of hyperandrogenism.

Women included in the study had withdrawal bleeding with progestin administration and normal levels of estradiol, prolactin, serum testosterone, free testosterone, and dehydroepiandrosterone sulphate. Additional criteria included a body mass index greater than 20.5 and no external appearances of PCOS such as hirsutism, acne, or alopecia. Twenty-four women met the inclusion criteria and were randomly assigned to receive 1,500 mg of metformin per day or placebo. Participants were assessed for ovulation during the study, and various laboratory evaluations were performed before and at the end of the study. Ovulation was determined by serum progesterone levels greater than 7 ng per mL (22 nmol per L) on day 22 or 23 after any episode of spontaneous bleeding. Fifty women with normal ovulation also were included in the study as controls.

Compared with the control group, the treatment participants had higher serum insulin levels and lower insulin sensitivity, but had similar androgen levels. During the three-month trial, eight of the women who received metformin had normal menses, and 16 ovulations were induced—significantly more than in those receiving placebo. The success of ovulation in the metformin group was not associated with changes in androgen, insulin, or insulin sensitivity parameters.

The authors conclude that metformin may be of use in inducing ovulation in women who are anovulatory and do not have hyperandrogenism. This effect occurs regardless of whether androgen or insulin levels are lowered.