Secondary Logo

Journal Logo


Suicide in Pregnancy

Dell, Diana L. MD; O'Brien, Benjamin W. MD

Author Information
  • Free

Patients with bipolar disorders who have an unplanned pregnancy can pose significant clinical challenges for both obstetric and psychiatric treatment.


A 26-year-old primigravida at 12 weeks' gestation presented to a psychiatric crisis center accompanied by her mother. The patient had a history of bipolar disorder, was no longer receiving medication, and was not doing well. The patient reported that her pregnancy was unplanned and undesired. Her pregnancy had resulted in the tumultuous ending of her relationship with the father of the fetus. Termination of pregnancy was not an option she would consider. She was currently living in a charity-funded home for pregnant single mothers who planned to place their infants for adoption.

At 8 weeks' gestation, this patient sought obstetric care at her local health department. She was told to discontinue all psychiatric medications and to follow up with her own psychiatrist. The patient had been receiving oxcarbazepine (Trileptal [Novartis Pharmaceuticals, East Hanover, NJ]; 300 mg twice a day; pregnancy category C) and bupropion (Wellbutrin-SR [Glaxo-SmithKline, Research Triangle Park, NC]; 150 mg daily; pregnancy category B) for the preceding 5 months. Unfortunately, she was in transition between state-funded mental health agencies and did not have a psychiatrist who had followed her over time. One week after discontinuation of her psychiatric medications, the patient sought care at a psychiatric crisis unit. She felt depressed and requested that she reinitiate medication. She received low-dose olanzapine (Zyprexa [Eli Lilly & Company, Indianapolis, IN]; 2.5 mg daily; pregnancy category B) and was advised to follow up with her obstetrician and outpatient psychiatrist.

Over the next 2 weeks, she was assessed twice in the same psychiatric crisis center. Each time, she reported worsening of her depression, but she denied suicidal ideation. She had not sought more prenatal care, nor had she sought psychiatric care providers in the public sector. No change in her medications was made at these visits. She was encouraged to keep her outpatient appointments.

At 12 weeks' gestation, the patient was admitted for inpatient psychiatric care. Since last seen in the crisis unit, she had cut off her hair and given away her belongings. She said her situation was hopeless and now refused medication or treatment. On the day of admission, she had reportedly held a loaded gun to her chest, but she “did not have the energy to pull the trigger.” She had written a four-page suicide note and had called her mother in another city to say goodbye.

The patient had been frequently hospitalized in the past for psychiatric problems; such hospitalizations began at age 18. She had been admitted for episodes of major depression and mania. She had a history of previous suicide attempts, the last of which occurred 8 months before this admission. After an overdose, she required treatment in the intensive care unit for severe respiratory depression. She had a chronic and well-documented history of noncompliance with treatment and follow-up. No family history of mental illness was reported.

Socially, she was one semester away from graduating with a bachelor's degree. She had recently quit her job. She reported having a close relationship with her parents and sibling; she described a happy childhood free of physical, emotional, or sexual abuse. She had been in a short, emotionally abusive relationship with the father of the fetus, but she denied other significant relationships. She denied drug or alcohol use.

Her mental status examination revealed that she was alert and oriented to person, place, and time. Her eye contact was poor. She appeared thin and disheveled, with a short, noticeably uneven haircut. Her speech was limited; she preferred to communicate through writing because “talking takes too much effort.” Her affect was restricted and congruent with her mood, which she described as “depressed and hopeless.” Mild psychomotor retardation was noted. No thought process abnormality was noted. Suicidal ideation was present: She planned to shoot herself in the chest. No delusions or hallucinations were noted. She had limited insight and poor judgment. Her admission diagnosis was “bipolar I disorder, most recent episode depressed, severe.”

The patient was admitted to the crisis stabilization unit under suicide watch until she was committed by the court to inpatient treatment. Initial laboratory evaluation was normal (complete blood count, metabolic studies, and thyroid profile), and her urine drug screen was negative. A quantitative human chorionic gonadotropin level obtained at admission was 154,642 mIU/mL.

Once admitted to the inpatient psychiatric unit, she was compliant with treatment. Her olanzapine dosage was increased from 2.5 mg to 15 mg daily; bupropion was reinitiated at 150 mg daily; and treatment with a prenatal vitamin and flax seed oil (2 g twice daily) was initiated. With resumption of treatment for her bipolar disorder, her mood brightened, she became less isolated, and she began interacting appropriately with staff and peers. She seemed less hopeless about her future as she prepared to return to the pregnancy care home where she had been living. After 3 days in the inpatient unit, she denied ongoing suicidal ideation and was released to outpatient psychiatric care and obstetric follow-up. Two days after discharge, the patient shot herself in the left chest; both she and the fetus died.


Obstetric care for women with bipolar disorder may represent the most intense clash between care for the mother and concern about the fetus. Many of the drugs used as mood stabilizers (especially the antiseizure medications, such as oxcarbazepine, used in this case) have been associated with fetal anomalies, although the risk is much lower than assumed by medical practitioners and the general public. In an academic practice that specializes in the care of women with psychiatric illness, it is not uncommon to hear that women who call to schedule an initial obstetric evaluation may be told to stop taking their psychiatric medications before seeing the doctor and before an appropriate risk–benefit analysis has been performed. This case is presented to focus attention on the need for an individualized evaluation of each woman who is receiving psychiatric medications, so that appropriate clinical judgment can weigh evidence for teratogenicity against the risk of not treating the psychiatric disorder.

The patient in this case report carried a long-standing diagnosis of bipolar affective disorder (also known as manic depression). The bipolar disorders actually represent a cluster of disorders of varying severity (bipolar I, bipolar II, cyclothymia, and bipolar disorder not otherwise specified). On the basis of this patient's history of hospitalization for at least one episode of mania, she would be placed in the most severe category, bipolar I.

Bipolar I disorder is a recurrent condition, with more than 90% of individuals who have one episode of mania experiencing many episodes. Manic episodes usually occur just before or just after an episode of major depression in a characteristic pattern for a given individual. The interval between episodes tends to decrease over time, making the overall impact of the disorder worse over time. Completed suicide occurs in 10% to 15% of individuals with bipolar I disorder. Lifetime prevalence in US community samples varies from 0.4% to 1.6%, with cases equally common in women and men (unlike bipolar II and major depression, which are more common in women).1

Bipolar disorder has an early age at onset in both women and men, with approximately 60% experiencing initial onset in adolescence or early adulthood (average age at onset is 21 years). Women with bipolar disorders usually have more depressive episodes than manic episodes. Women are also at greater risk for rapid cycling (four or more mood episodes during previous 12 months) and mixed episodes (rapidly alternating sadness, irritability, and euphoria)—both of which may increase the risk of suicide.2

Although women with bipolar disorders are at great risk in the postpartum period, pregnancy itself does not appear to have a major impact on the natural course of bipolar illness. Treatment of pregnant women with bipolar disorder, however, poses several difficult clinical dilemmas.

The most commonly used antimanic agents (lithium, antiseizure medications) all carry some fetal teratogenic risk in early pregnancy, as well as potential adverse effects in late pregnancy, labor, and delivery.3 Our patient received an antiseizure medication as a mood stabilizer. It was apparently working for her, but her caregivers feared it might be associated with a higher rate of fetal anomalies, so it was discontinued. Ideally, under these circumstances, there would be a discussion of the risk–benefit ratio while the patient is still euthymic, and a plan for treating relapses would be made.

Abrupt cessation of treatment with mood stabilizers is associated with a higher risk of early relapse and may contribute to a higher suicide risk.3 Abruptly stopping the mood stabilizer in this case may have contributed to early relapse because the patient presented with recurrent symptoms within 1 week. In general, the antiseizure medications have two windows of teratogenic risk. The first window of risk is in very early pregnancy, when the neural tube is forming; the second is near the end of the first trimester, during fusion of the midline facial structures. When this patient presented for care at 8 weeks' gestation, she was past the point of neural tube formation and not yet into the second window of risk, so a slower taper might have been possible.

When this patient presented with recurrent symptoms and the decision was made to resume medication, she and her practitioners had several therapeutic choices. In general, strategies for managing risk consist of withholding medications only during critical windows of risk, using hospitalization where needed for safety or structure, and considering electroconvulsive therapy if needed.4 Some practitioners would elect to limit the number of drugs to which the fetus was exposed and reinitiate the initial mood stabilizer; ultrasound assessment for normal development of spine and palate would be appropriate in the second trimester. Others would initiate a new mood stabilizer (lithium or atypical antipsychotic agent), which would represent a second fetal exposure but would bypass some of the fetal risk that had previously been of concern. In this case, lithium might have been a good choice because lithium's maximum teratogenic risk occurs in the first month of pregnancy (cardiac development) and clinically has been associated with a lower suicide risk. Instead, an atypical antipsychotic agent (olanzapine) was chosen, which was used at a very low dose. Although the atypical agents are widely and effectively used as mood stabilizers, they are relatively new drugs, with limited data regarding their impact on fetal development and subsequent behavior (although the limited data appear reassuring).5

Many psychiatric and obstetric practitioners are unaware of risk–benefit models for decision making about the use of psychotropic medication during pregnancy, leaving pregnant women vulnerable to recurrences of severe psychiatric illness.3 Such models have been proposed for depression and can be adapted for bipolar depression.6 For this patient, her history with her disorder and with its treatment are relevant to the treatment discussion; input from family members is often helpful. In this case, her mother felt strongly that her daughter needed to be back on medication (“We can't worry about the risk to the baby. Without treatment, they will both be dead”).

Pregnant women appear to have a significantly lower risk of suicide than women of similar age who are not pregnant, although that changes rapidly in the postpartum period, where suicide accounts for up to 20% of deaths. The lower risk of suicide during pregnancy appears to be a cross-national phenomenon, although the world literature on the subject is small. The reasons for these lower rates are unknown, especially because rates of depression during pregnancy are no lower than the general population—and in some populations may even be higher.7

Specific risk factors for suicide, starting in adolescence, include previous suicide attempts, major depression, substance abuse, living alone, and a history of physical or sexual abuse. Less specific but alarming factors include a recent dramatic personality change; intense psychosocial stressors; writing, thinking, and talking about death or dying; and altered mental status. The best predictor of suicidal behavior is a history of suicide attempt and current suicidal ideation. Suicidal ideation that includes a plan for suicide or evidence that the individual has been engaging in preparations for a suicide attempt are serious signs of short-term risk. In at-risk patients, the availability of a means for suicide should be reduced and, in particular, guns should be removed from the house.8

In this case, the patient's risk factors included the following: severe depressive symptoms; intense psychosocial stressors; writing, thinking, and talking about suicide; and altered mental status. Evidence that she was preparing for suicide included quitting her job, giving away her possessions, and saying goodbye to friends and family. Additionally, she had access to a gun.

It is not clear what precipitated this patient's return to suicidal ideation after discharge. Often, the safety and structure of an inpatient setting can make a patient appear more stable than she really is. It is also possible that this patient, who had undergone many previous hospitalizations, was able to feign improvement in an effort to return to her previous suicide plan. What is clear in this case is that the alteration of her psychotropic medication out of concern for fetal well-being probably contributed to her destabilization and to a fatal outcome.


1. Diagnostic and statistical manual of mental disorders. 4th rev ed. Washington, DC: American Psychiatric Association, 2000.
2. Amsterdam JD, Brunswick DJ, O'Reardon J. Bipolar disorder in women. Psychiatr Ann 2002;32:397–404.
3. Viguera AC, Nonacs R, Cohen LS, Tondo L, Murray A, Baldessarini RJ. Risk of recurrent bipolar disorder in pregnant and nonpregnant women after discontinuing lithium therapy. Am J Psychiatry 2000;157:179–84.
4. Miller LJ. Use of electroconvulsive therapy during pregnancy. Obstet Gynecol Surv 1995;50:10–1.
5. American Academy of Pediatrics, Committee on Drugs. Use of psychoactive medication during pregnancy and possible effects on the fetus and newborn. Pediatrics 2000;105:880–7.
6. Wisner KL, Zarin DA, Holmboe ES, Appelbaum PS, Gelenberg AJ, Leonard HL, et al. Risk–benefit decision making for treatment of depression during pregnancy. Am J Psychiatry 2000;157:1933–40.
7. Marzuk PM, Tardiff K, Leon AC, Hirsch CS, Portera L, Hartwell N, et al. Lower risk of suicide during pregnancy. Am J Psychiatry 1997;154:122–3.
8. Mann JJ. A current perspective of suicide and attempted suicide. Ann Intern Med 2002;136:302–11.
© 2003 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.