Bipolar disorder presents special challenges to women of childbearing age as well as to their families and clinicians. Problems include lower fertility rates, strong genetic loading, and potential fetal teratogenic risks as well as high risks of illness recurrence if treatment is discontinued abruptly (1–10). It is noteworthy, however, that family planning issues for women with bipolar disorder have received scant research consideration (1, 4). Pregnancy poses several clinical dilemmas, and evidence-based guidelines for the clinical care of this population remain very sparse. Extensive clinical experience suggests that women with bipolar disorder are often counseled to avoid or terminate pregnancy in order to avoid risks of potential fetal exposure to psychiatric medications or risk of recurrent illness.

Mood stabilizers, including valproate, carbamazepine, and lithium, are associated with teratogenic risks (2, 7). First-trimester exposure to lithium probably increases the risk of cardiac malformations, notably Ebstein’s anomaly, by several-fold, from a baseline risk of 0.005% (1 in 20,000 live births) to a risk ranging from 0.05% to 0.1% (1 in 2,000 to 1 in 1,000 live births) (2). Compared with lithium, anticonvulsants such as carbamazepine and valproic acid may pose even greater risks, including high rates (1%–5%) of neural-tube defects such as spina bifida as well as craniofacial anomalies, cardiac anomalies, microcephaly, and growth retardation (3, 10). Reproductive safety information about other, newer agents used to treat bipolar disorder remains very limited, leaving lithium as a plausible first-line option, especially during mid-to-late pregnancy (3, 10). Concern about teratogenic risks associated with the standard mood stabilizers can lead to incomplete consideration of the major risks associated with recurrences of bipolar disorder illness during pregnancy. These risks include not only the particularly high risk of early relapse after interruption of ongoing treatment but also the higher risk for postpartum recurrence as well as the impact of untreated psychiatric illness on the development of the fetus (1, 3, 6, 7, 10).

Although there are no empirically based treatment guidelines for the management of bipolar disorder during pregnancy, substantial progress has been made with improved information on the reproductive safety of psychotropic drugs used to treat bipolar disorder and a better understanding of the course of the disorder and the risks of recurrence during pregnancy and the postpartum period. Increasingly, women with bipolar disorder who wish to conceive seek preconception consultation to better understand the risks and benefits of treatment options (10). This report describes the family planning decisions made by patients with bipolar disorder after preconception consultation by a reproductive psychiatry subspecialty service in a major medical center.

A questionnaire was sent to women who sought outpatient consultation at the Perinatal and Reproductive Psychiatry Program at Massachusetts General Hospital during 1997–2000. The survey was designed to ascertain information regarding reproductive decisions that followed consultation designed to provide reliable information about the spectrum of risks associated with the maintenance or discontinuation of pharmacologic treatment, thus facilitating informed reproductive decisions.

The Perinatal and Reproductive Psychiatry Program was established in 1987 as a consultation service that provides specialized care to women suffering from reproductive-associated mood disorders, including affective disorders in pregnancy and the postpartum period.

The goal of the preconception consultation is to provide up-to-date information regarding the spectrum of risks associated with either maintenance or discontinuation of treatment with psychiatric medications. The main objective of the consultation is not to dictate treatment but to provide accurate information that patients may use to make personal decisions regarding treatment during pregnancy. In our clinical experience, patients with similar illness histories who are presented with the same risk-benefit information may make different decisions about maintenance or discontinuation of treatment in pregnancy. Although patients surveyed in this study were seen by at least four different clinicians in our program, the variability in treatment practices is minimal since there is general consensus among those who practice within the program regarding the management of bipolar disorder in pregnancy. A review of our program’s proposed tentative guidelines for the clinical management of bipolar disorder throughout pregnancy has been published recently (10).

A 13-question instrument addressed 1) demographic information, 2) reproductive history, 3) family planning practices, 4) reasons for seeking the original consultation, and 5) clinical outcome after consultation. Questions 3–5 were multiple choice, and multiple answers were encouraged. Patients were given a nominal payment to encourage cooperation. Massachusetts General Hospital’s institutional review board approved all study procedures, and subjects provided informed consent for use of questionnaire data for reporting in anonymous aggregate analyses.

Of 116 questionnaires sent, 13 were undeliverable, and 70 of 103 (68%) were completed and analyzed. The mean current age of the 70 respondents was 35.4 years (SD=5.7, range=20–49). Most were Caucasian (97%), were married (87%), and had a college degree (69%). The mean age at menarche was 13.2 years (SD=1.8, range=10–20). Irregular menstrual cycles were reported by half of the group, 11% (N=8) had undertaken infertility treatment, 19% (N=13) had at least one miscarriage, and 20% (N=14) had at least one therapeutic abortion. Before consultation, pregnancies per subject averaged 1.6 (SD=1.3, range=0–5), with a median of one live birth per subject.

Before consultation, 45% of the respondents (29 of 65) had been advised not to become pregnant by a health professional: 69% (20 of 29) by psychiatrists or other mental health professionals and 14% (four of 29) by primary care physicians or obstetricians. A spouse had urged 21% of the patients (six of 29) to avoid pregnancy, and 45% (13 of 29) reported that a parent or sibling had advised against pregnancy.

Respondents cited several reasons for seeking specialized consultation regarding management of their illness during a potential pregnancy. A majority, 52% (36 of 69), had been encouraged by a medical professional to seek such consultation, and 42% (29 of 69) reported seeking "a second opinion" on their own. Of those in our group, 55% (38 of 69) had been considering becoming pregnant and had sought information about the likely course of their illness and about the relative risks of the various treatment options, and 22% (15 of 69) had been pregnant at the time of consultation. About 25% of the group (17 of 69) sought consultation because they had previously experienced recurrences of bipolar disorder during pregnancy or the postpartum period.

Most respondents, 85% (55 of 65), reported that they had followed the treatment options outlined in their consultation. After consultation, 63% of the group (29 of 46) attempted to conceive on the basis of their personal assessment of the risks and benefits provided at the consultation and after review of these with their treating psychiatrist. Of those who tried to conceive, 69% (20 of 29) became pregnant within 12 months. The other 37% (17 of 46) chose to avoid pregnancy, including one who sought to adopt a child. The most commonly reported reasons to avoid pregnancy were fear of adverse effects of medicines on fetal development (56%, 10 of 18) and fear of illness recurrence if maintenance treatment were discontinued (50%, nine of 18). Fewer women expressed concerns about potential genetic transmission of bipolar disorder to offspring (22%, four of 18), reluctance to repeat previous pregnancy-associated illness (17%, three of 18), and fear that recurring mania or depression would adversely affect a fetus or existing children (17%, three of 18).

Similar proportions of women perceived that pregnancy had a positive influence on their illness course and overall well-being (47%, 16 of 34) as those who reported negative effects (53%, 18 of 34). In addition, one-half reported that becoming a mother had bolstered their self-esteem.

This follow-up survey of women with bipolar disorder who had been evaluated in consultation within our program has clear limitations, including incomplete surveying and bias toward women who were well educated, economically advantaged, and highly motivated to seek expert advice. Nevertheless, it yielded interesting preliminary insights into the concerns of women of childbearing potential who suffer from bipolar disorder. Approximately one-half of the 70 respondents had been advised against pregnancy by a psychiatrist, primary care physician, obstetrician, or family members, suggesting widespread bias against pregnancy for such women.

Our experience indicates that many women with bipolar disorder, regardless of educational and socioeconomic background, as well as physicians who care for them, are ill-informed about the relative risks of perinatal exposure to psychotropics and the high rates of relapse during pregnancy and the postpartum period without treatment (3, 9, 10). An important finding was that 37% of the patients chose not to pursue pregnancy when presented with very similar risk-benefit information on 1) the course and risk of recurrence of illness during pregnancy and the postpartum period and 2) the reproductive safety data of the various mood stabilizers, as compared to the 63% who attempted to conceive. This finding underscores the role of patient autonomy in clinical decision making and the importance of providing information about competing risks and potential benefits involved so that patients can make informed decisions about pregnancy.

A majority of respondents cited concerns about teratogenic risks as well as risk of recurrence after discontinuing maintenance medication as reasons for deciding against pregnancy. For this study group, both types of risk were given similar weight, but this outcome may reflect the impact of the consultation process itself and recent emphasis on maternal risk of treatment discontinuation in our center (10). On the basis of our clinical experience, fears about potential teratogenic risks of drug treatment during pregnancy appear still to have a strong restraining effect on both patients and physicians, despite the serious risks of treatment discontinuation in bipolar disorder, which have been appreciated more recently (6, 7, 10).

The study findings support our impression that providing accurate and balanced information about treatment options and relative risks, including the limits of current knowledge, can contribute importantly to informed family planning by women with bipolar disorder. We propose, specifically, that judgments concerning "reasonable risks" during pregnancy require shared responsibility but ultimately rest with the patient herself. Moreover, clinicians should resist automatic discontinuation of ongoing psychotropic medication in pregnancy without informing the patient of the considerable clinical risks involved (6, 7, 10) and taking her wishes about treatment and pregnancy into account. Studies in broader samples of women of childbearing age with bipolar disorder are required to clarify the unique reproductive health needs of this special and understudied population and to develop sound policies for their care.

Presented in part at the 154th annual meeting of the American Psychiatric Association, New Orleans, May 5–10, 2001. Received Jan. 25, 2002; revision received June 12, 2002; accepted June 26, 2002. From the Department of Psychiatry, Harvard Medical School, Boston; the Perinatal and Reproductive Psychiatry Program, Massachusetts General Hospital; the Department of Biostatistics, Harvard School of Public Health, Boston; and the International Consortium for Bipolar Disorder Research, McLean Hospital, Belmont, Mass. Address reprint requests to Dr. Viguera, Perinatal and Reproductive Psychiatry Program, WACC 812, Massachusetts General Hospital, 15 Parkman St., Boston, MA 02114; aviguera@partners.org (e-mail). Supported in part by NIMH grant MH-01609 and a Young Investigators Award from the National Alliance for Research on Schizophrenia and Depression (to Dr. Viguera), a grant from the Stanley Foundation (to Dr. Cohen), and an award from the Bruce J. Anderson Foundation and the McLean Hospital Psychopharmacology Research Fund (to Dr. Baldessarini).