Reproductive function and risk for PCOS in women treated for bipolar disorder
Natalie L Rasgon,Lori L Altshuler,Lynn Fairbanks,Shana Elman,Jose Bitran,Rodrigo Labarca,Mohammed Saad,Ralph Kupka,Willem A Nolen,Mark A Frye,Trisha Suppes,Susan L McElroy,Paul E Keck Jr,Gabriele Leverich,Heinz Grunze,Joerg Walden,Robert Post,Jim Mintz
DOI: https://doi.org/10.1111/j.1399-5618.2005.00201.x
Abstract:Introduction: This study examined the reproductive function and prevalence of polycystic ovary syndrome (PCOS) in women with bipolar disorder taking antimanic medications. Method: Women aged 18-45 treated for bipolar disorder and not taking steroid contraceptives were recruited to complete questionnaires about their menstrual cycle and to provide blood samples for measurement of a range of reproductive endocrine and metabolic hormone levels. Eighty women participated in completing the questionnaires and 72 of them provided blood samples. Results: Fifty-two of the 80 women (65%) reported current menstrual abnormalities, 40 of which (50%) reported one or more menstrual abnormalities that preceded the diagnosis of bipolar disorder. Fifteen women (38%) reported developing menstrual abnormalities since treatment for bipolar disorder, 14 of which developed abnormalities since treatment with valproate (p = 0.04). Of the 15 patients reporting menstrual abnormalities since starting medication, 12 (80%) reported changes in menstrual flow (heavy or prolonged bleeding) and five (33%) reported changes in cycle frequency. No significant differences were observed between women receiving or not receiving valproate in mean levels of free or total serum testosterone levels. This was true for the total sample and for the sub-group without preexisting menstrual problems. However, within the valproate group, duration of use was significantly correlated with free testosterone levels (r = 0.33, p = 0.02). Three of the 50 women (6%) taking VPA, and 0% of the 22 taking other antimanic medications, met criteria for PCOS (p = 0.20). Other reproductive and metabolic values outside the normal range across treatment groups included elevated 17 alpha-OH progesterone levels, luteinizing hormone: follicle-stimulating hormone ratios, homeostatic model assessment (HOMA) values, and low estrogen and dehydroepiandrosterone sulfate (DHEAS) levels. Preexisting menstrual abnormalities predicted higher levels of 17 alpha-OH progesterone, free testosterone, and estrone as well as development of new menstrual abnormalities. Body mass index (BMI) was significantly positively correlated with free testosterone levels and insulin resistance (HOMA) across all subjects, regardless of medication used. Conclusions: Rates of menstrual disturbances are high in women with bipolar disorder and, in many cases, precede the diagnosis and treatment for the disorder. Treatment with valproate additionally contributes significantly to the development of menstrual abnormalities and an increase in testosterone levels over time. A number of bipolar women, regardless of type of medication treatment received, have reproductive and metabolic hormonal abnormalities, yet the etiology of such abnormalities requires further study. Women with preexisting menstrual abnormalities may represent a group at risk for development of reproductive dysfunction while being treated for bipolar disorder.