Polycystic ovarian syndrome (PCOS) occurs in 4-7% of women and is characterized by irregular menstrual cycles and hyperandrogenism (facial hair, acne, male-pattern hair loss, acne, or elevated male hormone levels). The majority of women with PCOS also suffer from obesity and insulin resistance. PCOS has been associated with a spectrum of health problems including infertility, diabetes, and possibly heart disease and endometrial cancer. Recently there has been concern that women with bipolar disorder who are treated with the mood stabilizer valproate (VPA; brand name Depakote or Depakene), may be at higher risk for PCOS, although the data have been somewhat conflicting.
In a first study from CWMH researcher Dr. Hadine Joffe, 300 women with bipolar disorder (range 18-45 years) participating in the Systematic Treatment Enhancement for Bipolar Disorder (STEP-BD) study sites were evaluated for PCOS. This study included women with bipolar disorder (ages 18-45) who had received treatment with a mood stabilizer (valproate, carbamazepine, lithium, lamotrigine, topiramate, gabapentin, or oxcarbazepine) for at least 3 months. 230 women were evaluated for treatment-emergent PCOS, which included retrospective assessment of menstrual cycle patterns, menstrually timed assays of serum hormone levels, and screening for physical signs of hyperandrogenism (facial hair, acne, male-pattern hair loss). Menstrual irregularity with hyperandrogenism developed in 9 (10.5%) of 86 women on valproate compared to 2 (1.4%) of the 144 women on a non-valproate anticonvulsants or lithium. This represents a 7.5-fold increase in risk for PCOS among valproate users. Menstrual irregularity emerged early, developing within 3 months in half of the women. At higher risk for PCOS were women who started treatment with VPA at an earlier age and women. Ultrasound examination of the ovaries revealed that valproate use was not associated with typical polycystic ovarian morphology.
In the second study from the same team of researchers, a group of women with treatment-emergent PCOS were followed longitudinally, including seven women who developed PCOS on valproate. PCOS features resolved in three of the four women discontinuing treatment with valproate. The PCOS symptoms persisted in all 3 women continuing valproate. While menstrual cycle irregularities usually improved among the women who discontinued valproate, body weight and polycystic ovarian morphology remained the same.
The first study suggests that PCOS may be more common among bipolar women treated with valproate, occurring in approximately one out of every 10 women. Symptoms of PCOS tended to emerge early, often within the first three months of treatment. While PCOS may be common in this population, the second study suggests that treatment-emergent PCOS symptoms may be reversible upon discontinuation of valproate; however, this study was limited by its small sample size. Given the spectrum of health problems associated with PCOS and the unclear reversibility of this syndrome, the use of valproate in women with bipolar disorder must be considered carefully. Women treated with valproate must be informed of their risk for PCOS and should be monitored for signs of PCOS, especially during the first year of valproate use. Weight change, menstrual irregularity, excess facial hair, male-pattern hair loss and acne are signs than should alert the clinician to the possibility of treatment-emergent PCOS.
Ruta Nonacs, MD PhD
Joffe H, Cohen LS, Suppes T, et al. Longitudinal follow-up of reproductive and metabolic features of valproate-associated polycystic ovarian syndrome features: A preliminary report. Biol Psychiatry 2006; 60: 1378-1381.