Longitudinal evaluation of reproductive function in women treated for bipolar disorder

Background: We assessed reproductive endocrine and metabolic markers in women treated for bipolar disorder over a 2-year time period, controlling for valproate use. Methods: Twenty-five women ages 18-45 with bipolar disorder underwent longitudinal evaluations. Subjects completed a reproductive health questionnaire and endocrinological exam at baseline. Total and free testosterone, progesterone, LH, FSH, fasting insulin and glucose, and other hormones were measured across the menstrual cycle at baseline and at 2-year follow-up. Results: Ten subjects were currently receiving valproate as a mood stabilizing agent; of the remaining subjects, six received lithium and five received atypical antipsychotics. Of all subjects, 41.7% reported current oligomenorrhea, while 40% reported oligomenorrhea before starting medication. Rates of oligomenorrhea and clinical hyperandrogenism did not differ by medication use. Eighty percent of women had a high homeostatic model assessment of insulin resistance (HOMA-IR) at baseline; all other measures were normal. Over time, all subjects exhibited a significant decrease in luteal phase progesterone and increase in free testosterone concentrations. Valproate use was associated with an increase over time in total testosterone. Baseline values and changes in BMI were similar across groups. Limitations: Limitations include small sample size and the absence of a control group. Conclusion: We confirm our previous observations of high rates of menstrual abnormalities, hyperandrogenemia and insulin resistance in women with bipolar disorder. These results tentatively support the role of valproate in hyperandrogenemia; however, rates of oligomenorrhea and clinical hyperandrogenism did not differ between medication groups. (c) 2005 Elsevier B.V. All rights reserved.