Metformin as a Treatment for Antipsychotic Drug Side Effects: Special Focus on Women With Schizophrenia

Metformin is an antidiabetic drug that may have, as demonstrated by developing evidence, considerable therapeutic application for the treatment of antipsychotic drug side effects. Studies published over the past 6 years have indicated that metformin can attenuate or reverse some of the weight gain induced by antipsychotic drugs (1–3) and that this effect may be additive to lifestyle intervention (4). The weight-loss effect of adjunctive metformin may be stronger in first-episode or drug-naive patients started on second-generation antipsychotic medications, especially those that are associated with the greatest weight gain, such as clozapine and olanzapine (5–7). These studies have also shown that metformin concomitantly improved measures of glucose metabolism, such as reduction in fasting glucose and insulin levels and measures of insulin resistance, in the same antipsychotic-treated patients (4, 6, 7). Although metformin may decrease weight and improve glucose metabolism in the same patient, it is not clear how closely these effects are correlated or whether they operate through the same or different underlying mechanisms.

The article in this issue by Wu et al. (8) presents additional support for metformin's therapeutic effects on weight and metabolic parameters, but it also adds important new data on therapeutic effects of metformin on a side effect that is particularly important for women with schizophrenia treated with antipsychotics. Amenorrhea was reported in 9%–12% of first-episode schizophrenia patients in the European First-Episode Schizophrenia Trial (9), and menstrual dysfunction (irregular or absent menses) was noted in 50%–63% of female patients in a small study of chronic schizophrenia (10). It has been suggested that amenorrhea and menstrual irregularities may be more frequent occurrences in patients treated with prolactin-elevating antipsychotics, such as risperidone, sulpiride, and some first-generation antipsychotics, although some studies have not shown differential menstruation-related side effects in patients treated with prolactin-elevating compared with prolactin-sparing antipsychotics; antipsychotic-specific risk estimates have not been established.

On the basis of previous studies of the effectiveness of metformin in restoring menstruation in women with polycystic ovarian syndrome (11) and in nonpsychotic patients with amenorrhea (12), Wu and colleagues designed a double-blind study of metformin (1000 mg/day) or matched placebo in antipsychotic-treated schizophrenia patients with amenorrhea. The authors report that 66% of the metformin-treated patients, compared with only 2% of the placebo patients, resumed menstruation during the 6-month study, and the mean time for restoration of menstruation was 2 months. Metformin also decreased levels of prolactin, luteinizing hormone (LH), and follicle-stimulating hormone (FSH) and significantly decreased weight, insulin levels, and insulin resistance in these patients. Logistic regression analysis suggested that decreases in body weight, insulin resistance, and levels of prolactin, LH, and FSH were all related to the drug's therapeutic effect of restoring menstruation, although the specific contribution of these effects to the mechanism by which metformin restores menstruation is not clear. Many antipsychotic drugs exert some of their biological or clinical effects by interfering with dopamine transmission through blockade of postsynaptic dopamine receptors; this is believed to be an important mechanism through which these drugs raise prolactin and other hormone levels. Metformin has been shown to increase dopaminergic tone in patients with polycystic ovarian syndrome (13), and this may be one of its mechanisms of action in the clinical effects reported here. It is also important to note that there were no differences in side effects between the metformin and placebo groups, which indicates that metformin was a safe treatment as well as an efficacious one.

The participants in the Wu et al. study were first-episode schizophrenia patients with a mean duration of illness of less than 1 year and a mean duration of antipsychotic treatment of about 6 months. Whether a similar degree of effectiveness of metformin for correction of amenorrhea and gonadotropin hormone abnormalities will be found in chronic schizophrenia patients who have been treated with antipsychotics for many years will need confirmation by additional research. The effects of metformin in reducing weight in antipsychotic-treated schizophrenia patients have been stronger or more consistent in published studies of first-episode or newly treated schizophrenia than of chronic schizophrenia (3), although preliminary data from studies of long-term antipsychotic-treated schizophrenia presented at recent meetings suggest that metformin may also have weight-loss effects in patients with chronic schizophrenia.

These reports highlight the need for clinical psychiatrists to be more fully aware of, and to monitor more closely, the potential metabolic and endocrine effects of antipsychotics, as well as the developing treatment options supported by results from high-quality scientific trials such as the Wu et al. study (8). The results of this study also raise questions about whether clinical recommendations should be developed for more routine use of metformin in women treated with antipsychotics who develop irregular menses or amenorrhea and whether this should be a time-limited exposure or a longer-term adjunctive treatment. Similar questions have been raised about possible modification of criteria for more routine use of metformin for treating prediabetes in antipsychotic-treated schizophrenia patients (14) or long-term use of metformin for controlling weight gain in antipsychotic-treated patients. A recent report from the Diabetes Prevention Program (15) shows that long-term treatment (up to 10 years) with metformin was not associated with clinically significant adverse events other than self-reports of gastrointestinal symptoms and that long-term treatment was associated with significantly reduced body weight and waist circumference. Although longer-term studies of metformin effects in antipsychotic-treated patients may be needed to generate data on which to base definitive clinical recommendations, the effects of long-term use of metformin in the Diabetes Prevention Program studies are consistent with long-term therapeutic benefit and a quite favorable risk-benefit ratio.