Get access

Valproate, bipolar disorder and polycystic ovarian syndrome


Corresponding author: Roger S. McIntyre MD, FRCPC, University Health Network, Mood Disorders Psychopharmacology Unit, 399 Bathurst Street, ECW-3D-008, Toronto, Ontario, Canada M5T 2S8. Fax: 416-603-5368; e-mail:


Background:  Persons with bipolar disorder are often overweight and cluster risk factors for cardiovascular disease. Some antibipolar agents adversely impact upon weight and the lipid milieu. Recent data suggest that valproic acid, a commonly prescribed mood stabilizer, may be associated with polycystic ovarian syndrome (PCOS). This adverse event has not been systematically studied in bipolar disorder.

Method:  Thirty-eight female subjects, aged 18–50 years, meeting DSM-IV criteria for bipolar I or II disorder, in any phase of illness were evaluated. Eighteen females received valproate (sodium valproate and valproic acid) and 20 females received lithium. Patients completed questions regarding their menstrual, reproductive and medical histories. During the follicular phase they were assessed for weight, body mass index (BMI kg/m2), and changes in the reproductive endocrine milieu that included morning estradiol, progesterone, follicle-stimulating hormone (FSH), luteinizing hormone (LH), sex-hormone binding globulin (SHBG), androstenedione, dehydroepiandrosterone-sulfate (DHEAS), testosterone, free testosterone, prolactin and thyroid-stimulating hormone (TSH). The blood was also analyzed for fasting metabolic parameters which included total cholesterol (TC), high-density lipoprotein (HDL), low-density lipoprotein (LDL), insulin, glycosylated hemoglobin (HbA1C), insulin-like growth factor 1 (IGF-1), insulin-like growth factor binding-protein 1 (IGFBP-1), fasting blood glucose and morning leptin.

Results:  Nine (50%) of the valproate-treated females had menstrual abnormalities versus three (15%) of the lithium-treated females (p < 0.05). Valproate-treated females had significantly higher levels of follicular phase androgen concentrations than lithium-treated females (p < 0.05). Nine (50%) of females who were overweight (BMI ≥ 25 kg/m2) and with a history of menstrual irregularities also exhibited laboratory evidence of hyperandrogenism (p < 0.05). Persons receiving valproate exhibited significant increases in fasting biochemical parameters suggestive of an adverse metabolic syndrome (p < 0.05). Leptin levels were significantly elevated in the valproate-treated females (p < 0.05).

Conclusions:  In this pilot, open-label cross-sectional study, valproate-treated females exhibited higher rates of menstrual abnormalities and biochemical evidence of both hyperandrogenism and adverse metabolic parameters when compared with lithium-treated females. These preliminary data suggest that valproate may, in some predisposed females, adversely impact upon the reproductive endocrine milieu and result in aspects of the metabolic syndrome (1).