Current issues: women and bipolar disorder.

While the treatment of bipolar disorder (BD) is typically complex, the treatment of women with bipolar disorder is even more challenging because clinicians must also individualize treatment based on the potential for pregnancy, drug interactions with oral contraceptives, and an increased risk of endocrine diseases that can either impact the course of illness or become manifest with some treatments. Women with BD should be checked for hypothyroidism, and if prescribed antidepressants, carefully watched for rapid cycling or a mood switch to mania, hypomania, or a mixed state. Several medications interact with oral contraceptives or increase the risk of developing polycystic ovary syndrome. Consideration of possible pregnancy is essential, and should be planned in advance whenever possible. Rates of recurrence have been shown to be equal in pregnant and nonpregnant women with BD. Risks of medication to the fetus at various points of development must be balanced against the risks of not treating, which is also detrimental to both fetus and mother. The postpartum period is a time of especially high risk; as many as 40% to 67% of women with BD report experiencing a postpartum mania or depression. The decision to breastfeed must also take into account the adverse impact of sleep deprivation in triggering mood episodes. In order to best address these issues, clinicians must be familiar with the data and collaborate with the patient to assess risks and benefits for the individual women and her family.

Bipolar disorder (manic depressive illness) affects approximately 1% to 2% of the population, and as many as 3.9% when bipolar I disorder and bipolar II disorder are aggregated.[1-3] While the prevalence of bipolar disorder (BD) is comparable in men and women, there are several aspects of bipolar disorder that require unique consideration in women. This manuscript reviews the course of illness considerations for women with bipolar disorder, how bipolar disorder impacts reproductive function in women, and considerations for the treatment of women who are planning pregnancy, or who are pregnant, postpartum, and/or breastfeeding.

The impact of gender on course of illness of bipolar disorder

There are few clinical characteristics that reliably differentiate men and women with bipolar disorder. Multiple authors have reported that women experience more depressive episodes over the course of their illness compared with men.[4-6] However, the concern that women may be more willing to report a prior depressive episode has not received adequate attention. It is also reported that women with bipolar disorder are more likely to experience rapid cycling,[6-8] mixed mania,[9-12] and antidepressant-induced manias[13] compared with men with bipolar disorder. Burt and Rasgon[14] point out that this difference may be due to inadequate mood stabilization and excessive use of antidepressants in women. Recent randomized evidence suggests that antidepressants added to adequate doses of antimanic medications do not improve outcomes in bipolar depression.[15] Taken together, at this juncture, when a woman with bipolar disorder presents with depression or rapid cycling, it. appears prudent to optimize mood stabilizers, check for hypothyroidism (which is more common in women), and judiciously reevaluate the use of antidepressant medications.

The impact of menses and menopause on the course of illness of women with bipolar disorder

Evidence on the impact of the menstrual cycle on course of illness of bipolar disorder remains mixed. Some studies report that women with bipolar disorder report frequent premenstrual mood disturbances,[16-17] while other studies report mixed findings.[13,18] Little is known about the influence of menopause on bipolar disorder in women. Various reports suggest that, menopause can improve, worsen, or not impact the course of mood symptoms in women with bipolar disorder.[19] Blehar et al[16] found that as many as 20% of postmenopausal women with bipolar disorder reported severe emotional disturbances during the menopausal transition. Some researchers have described this as a conversion to a rapid cycling variant of bipolar disorder.[20] .More data is needed to understand whether these hormonal transitions directly impact the course of bipolar illness. Careful evaluation of individual women with respect to menses and menopausal status appears warranted, with the institution of symptomatic treatment, if needed.

The impact of pregnancy and the postpartum period on the course of illness of bipolar disorder

A common misconception is that pregnancy is protective against psychiatric symptoms. Data clearly indicate that this is not the case in bipolar disorder, where rates or recurrence have been shown to be equal in pregnant and nonpregnant women.[21] Factors associated with a higher risk of relapse during pregnancy include abrupt discontinuation of mood stabilizers, a history of four or more prior mood episodes, and prior intrapartum mood episode(s).[21]

The postpartum period is a time of particularly high risk for women. While estimates vary, the risk for relapse during the puerperium range from 20% to 50%. [22-24] As many as 40%o to 67% of women with BD report experiencing a postpartum mania or depression.[25-26] Women with BD have a 100-fold higher risk than women without a psychiatric illness history of experiencing postpartum psychosis. This form of psychosis usually starts within 3 weeks of childbirth, and the initial presentation is often delusions.[27-28] It is important to differentiate postpartum depression from the “baby blues.” Baby blues consist of mood lability and depressed mood and occurs during the first 2 weeks postpartum, but does not persist and is not associated with delusions or marked impairment of functioning.

Medication effects in women with bipolar disorder

Differences in treatment response

Currently, little is known about what medications might be associated with most, benefit, in women with bipolar disorder. At present, only lithium response has been well studied. In a meta-analysis of 17 lithium treatment studies, there was no gender difference in response to lithium observed.[29] The following sections will address other medication effects that may impact choice of treatment for women with bipolar disorder.

Menstrual and ovarian changes related to medication treatments

Polycystic ovary syndrome (PCOS) is a serious endocrine disorder that affects women in their reproductive years.[30-32] PCOS is a syndrome defined by the presence of: (i) ovulatory dysfunction (ie, polymenorrhea, oligomenorrhea, or amenorrhea); (ii) clinical evidence of hypcrandrogenism or hyperandrogenemia; and (iii) exclusion of other endocrinopathies (eg, hyperprolactinemia, thyroid dysfunction, adrenal hyperplasia, or Gushing syndrome).[33] PCOS is the leading cause of anovulatory infertility and hirsutism, and is associated with multiple reproductive, metabolic disorders, and general health disorders, including increased risk of miscarriage, insulin resistance, hyperlipidcmia, cardiovascular disease, and endometrial cancer. In the general population of reproductive-age women, the prevalence of PCOS is estimated to be between 4% and 7%, but may be as high as 11 %.[32,34]

An association between the development of PCOS and the use of antiepileptic drugs (AEDs) was first suggested by Isojarvi et al.[35] This study looked at the prevalence of PCOS spectrum symptoms in 238 women with epilepsy, and suggested that use of valproate is associated with an increase in reproductive disorders. In a follow-up study,[36] 12 women with epilepsy were switched from valproate to lamotrigine, to assess whether changes in body mass index, insulin levels, and associated other symptoms were reversible. Twelve months after switching, the 12 women had lost weight and exhibited decreased BMI, insulin, and testosterone levels. The number of women with polycystic ovaries decreased from 11 to 7, and the number with menstrual abnormalities decreased from 7 to 2.

While these findings raised concern for the use of valproate in women, the studies were all conducted in women with epilepsy, and it was unclear if the association would be present, in other groups. Subsequent, studies assessed the relationship of valproate use and risk for PCOS in women with bipolar disorder. Rasgon et al[37] conducted a small pilot study in 22 women with bipolar disorder, receiving lithium monotherapy, valproate monotherapy, or lithium-valproate combination therapy. None of the patients in the study met criteria for PCOS, and there was no relationship between valproate or lithium therapy and PCOS. She followed this with a larger cross-sectional trial including 96 women, aged 1 8 to 45, who were being treated for a DSM-IV diagnosis of bipolar disorder I, II or NOS, and who had received longterm treatment with an antimanic agent, through the Stanley Foundation Treatment Network.[38] Of the 80 women with complete questionnaire data, 52 (65%) reported current menstrual abnormalities. While only 15 women (38%) reported new menstrual abnormalities

since treatment for bipolar disorder, 14 of these occurred since treatment with valproate (P=0.04). No significant differences were observed between women receiving or not receiving valproate in mean levels of free or total serum testosterone levels (n=72). Of the 50 women taking VPA, 3 (6%) met criteria for PCOS< compared with 0% of the 22 women taking other antimanic medications (

Another small study included 38 women with bipolar disorder, receiving valproate or lithium monotherapy for at least 2 years.[39] Menstrual irregularities were reported by 50% of the valproate-treated patients and 15% of the lithium-treated patients. Free testosterone and androstenedione levels were significantly higher than the reference range in valproate-treated patients, and LH was elevated in both treatment groups. The investigators concluded that valproate may result, in some aspects of the metabolic syndrome in some women with bipolar disorder. This study is limited by its small size and lack of a control group.

Joffe et al[40] examined 300 women with bipolar disorder, between the ages of 18 and 45, participating in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). Medication and menstrual-cycle histories were obtained, and hyperandrogenism was assessed. Among 230 women with complete assessments, oligomenorrhea with hyperandrogenism developed in 9 of 86 (10.5%) women on valproate and in 2 of 144 (1.4%) women on an alternative anticonvulsant or lithium (relative risk 7.5, 95% CI 1.7-34.1, P=0.002). When it occurred, oligomenorrhea began within the first 12 months of valproate use. This study demonstrated an association between valproate and new-onset oligomenorrhea with hyperandrogenism in women with bipolar disorder. A subsequent follow-up study completed follow-up assessments (after 17+I-7-months) in 14 women (5/9 with treatment-emergent PCOS, 9/1 9 valproate use "6 months).[41] Of 7 women who developed valproate-asso ciatcd PCOS, reproductive features of PCOS remitted in 3/4 women discontinuing valproate and persisted in all 3 continuing valproate. Compared with women continuing valproate, menstrual-cycle irregularities improved among valproate discontinucrs whose PCOS features remitted (P=0.01). There was a trend toward lower serum testosterone (P=0.06). Body weight, was unchanged.

Valproate may also be associated with PCOS features because increase in body weight or insulin resistance secondary to valproate therapy[36,42-43] may lead to the development of PCOS through insulin effects in the ovary.[44] However, menstrual-cycle irregularities or PCOS are uncommon in women with obesity or type 2 diabetes.[45-47] Prospective research is needed to examine the relationship between weight, insulin resistance, and predisposition or development of PCOS features.

The collective literature demonstrates that rates of menstrual disturbances are high in women with bipolar disorder, regardless of their treatment history. It appears that treatment with valproate further predicts the development of menstrual abnormalities and an increase in testosterone levels over time. However, little is known about the additive impact of previous exposure, duration of exposure, and age of women who are most vulnerable to development of this constellation of symptoms.[48] More research is needed to understand the relationship between etiology of reproductive and hormonal irregularities, onset of bipolar disorder, and treatment history.

Endocrine effects of medication treatments

Women are at greater risk than men for the development of lithium-associated hypothyroidism. Clinical hypothyroidism during lithium treatment is present, in 14% of women, versus 5.5% of men.[49] Lithium-treated women may also be at higher risk for lithium-induced thyroiditis.[13]

Effects of pharmacotherapy on oral contraceptives

The efficacy of oral contraception (OC) can be impaired by concomitant use of medications that induce liver enzymes (eg, carbamazepine, oxcarbazepine), which may be secondary to enhanced hepatic metabolism of the OC hormones. Therefore, if women are prescribed these medications for treatment of symptoms of bipolar disorder, clinicians should advise them to use barrier methods of birth control, monitor for spotting, and/or work with the gynecologist to increase oral contraceptive pill (OCP) dose. Conversely, OCPs induce lamotrigine metabolism, such that increased lamotrigine doses are often required for women on OCPs. While no drug-drug interactions with OC have been reported to date with valproate, lithium, or the atypical antipsychotics, further study is required in women with bipolar disorder.

In summary, there is no systematic controlled data to demonstrate that certain treatments or more effective for men and women. Instead, providers should carefully weigh potential side effects and interactions associated with treatments, and the importance of those risks for individual women.

Treatment of bipolar disorder during pregnancy and postpartum

Medication use during pregnancy

We strongly recommend that clinicians discuss plans for conception with all women with bipolar disorders who arc of childbearing potential. Recent work suggests that when women with bipolar disorder are provided accurate and balanced information about the potential risks and benefits they face, 37% choose not to pursue pregnancy.[50] Prenatal counseling should include discussion of possible risks of taking medications during pregnancy, risks to the patient, and child of escalating or uncontrolled symptoms of bipolar disorder, and the risk of genetic transmission of bipolar disorder to the child.[14]

In bipolar women who are pregnant, the use of medications must, be assessed in terms of adverse fetal or neonatal effects, in addition to the usual concerns for effectiveness, tolcrability, and safety for the mother. Before pregnancy begins, the patient, family, and clinician should detailed a plan of potential interventions in case of recurrences or exacerbations of mood episodes. For example, deciding if electroconvulsive therapy, which is relatively safe in pregnancy, would be the first choice if a severe depressive episode occurred.

One of the most difficult problems for women with bipolar disorder is the lack of effective nontcratogenic treatments. First-trimester exposure to the traditional mood stabilizers (lithium, valproate, and carbamazepine) is associated with an increased risk of fetal malformations.[51-53] Given this risk, many women with bipolar disorder choose to discontinue medications during pregnancy and sometimes, while trying to conceive. When this is done abruptly, women are at increased risk for relapse. Viguera et al[21] reported recurrence rates following lithium discontinuation in a cohort, of 101 pregnant and nonpregnant women. Over the 64-week period following lithium discontinuation, recurrences occurred in 85.7% of the prcgnant/postpartum women and 67.8% of the nonpregnant women. Recurrence rates were less when lithium was discontinued via a gradual taper (15 to 30 days). In some cases, it is preferable to continue the medication while carefully monitoring fetal development with high-resolution ultrasound.

Once the high risk associated with first-trimester exposure to certain medications has passed, many women who had discontinued medication then consider restarting pharmacotherapy. There is no clear guidance as to whether this should be done automatically, or whether it is preferable to wait for early signs of potential relapse. In women with a history of self-harm or protracted recovery from previous episodes, impaired insight, or a strained support, system, reinstatement of pharmacological treatment may reduce overall risk to both mother and fetus.[54]

Several careful reviews of medications useful in bipolar disorder and their implications for pregnancy and postpartum use are available[14,54-58] An overview of safety concerns for commonly used treatments is presented in Table I. It. should be noted that much of what, is known about the safety parameters of anticonvulsant drugs come from registries of epileptic women, and safety information for other classes of drugs also stem from diverse diagnostic subgroups.

General recommendations for treatment of bipolar women during pregnancy include minimizing pharmacotherapy if clinically feasible, particularly during the first trimester.[14] If treatment is initiated or continued, use of monotherapy at the minimal effective dose is recommended. Given that lamotrigine monotherapy has the largest safety database, Gentile[56] suggests this agent as the first-line mood stabilizer during pregnancy. In pregnancies with risk for neural tube deficits, folate is prescribed at 4 mg/day, compared with 0.4 mg. prescribed in lower-risk pregnancies.

Treatment during the postpartum period, and while breastfeeding

Regardless of whether a mood episode occurs during pregnancy, the postpartum period is associated with particularly high risk for relapse. Women with bipolar disorder have a 100-fold higher risk than women without a history of psychiatric illness of developing a postpartum psychosis.[27] Some suggest beginning prophylaxis in the second or third trimester of pregnancy, when there is less teratogenic risk[59,60] although there is no consensus on when to begin prophylaxis.[61] During pregnancy, the patient, and her doctor must make plans for the postpartum period, including discussion of options for prophylaxis.

Several studies have suggested the positive benefits of prophylaxis. A small open-label study of women at risk for puerperal psychosis (women with bipolar disorder diagnoses, or previous episodes of postpartum psychosis) added lithium prophylaxis in the third trimester of pregnancy or immediately after delivery. Of 21 women observed, only 2 had a recurrence of their psychotic illness.[60] Another small study included 27 women with bipolar disorder.[62] Only 1 of the 14 patients starting prophylactic agents during the postdelivery period relapsed within the first 3 postpartum months, while 8 of the 13 who did not receive prophylaxis showed evidence of recurrent mood instability during those 3 months. Similar positive benefits were observed in other small studies.[63-64]

In contrast, Wisner et al[65] offered divalproex plus symptom monitoring, or symptom monitoring alone, for immediate postpartum management of 26 women with bipolar disorder. There were no significant differences between groups in the proportions of women who developed postpartum mood episodes over the 20-week observation period. The time to development of a mood episode also did not. vary between groups.

Treatment decisions about, medication use postpartum should be based on the mother's clinical status and previous course, regardless of breastfeeding status.[14] In other words, the mother's health and stability should take priority over the feeding method of the infant. While breastfeeding is associated with many potential benefits to both mother and child, the sleep disruption associated with being the sole source of food for a newborn is contraindicated for many bipolar women.[55] Women should explore options to ensure adequate sleep, including arranging for other adults to feed the infant, and expressing milk earlier in the day for night feedings. The mother and her partner should be educated about the possible risks of breastfeeding while taking medication, and the infant should be monitored as needed. Again, monotherapy with the lowest possible dose of medication is the preferred treatment option, if pharmacotherapy is pursued.

Nonpharmaeological treatment options during pregnancy and lactation

Because of concerns over the use of traditional medications during pregnancy, there has been great interest in exploring the utility of omega-3 fatty acids for women planning pregnancy, pregnant, or lactating. Unlike traditional treatments, addition of omcga-3 fatty acids may benefit both mother and fetus, as adequate intake of omega-3 fatty acids is necessary for optimal fetal and infant brain and nervous system development, and (DHA) is selectively transferred to the developing fetus during pregnancy.[66-73] Stores of eicosopentaenoic acid (EPA) are progressively depleted during pregnancy.[74] Hibbeln and Salem[75] have hypothesized that this may predispose women to affective episodes. Additionally, research suggests that, pregnant women only achieve 20% to 60% of recommended omega-3 fatty acid intake.[76] Omega-3 fatty acids (DHA + F,PA) have been administered to pregnant women with various other disorders, without adverse effects.[7779]

A small randomized placebo-controlled study assessed the benefit of an omega-3 fatty acid (DHA) in women planning pregnancy.[80] This study also incorporated a brief psychosocial educational intervention, involving the woman and close supporters. The 10 participants tolerated the trial well, with no serious adverse events reported. Two of the women in the active group completed the 52-week trial (33.3%), and of those with premature discontinuation, 3 were due to emerging or worsening mood symptoms (50%) and 1 due to noncompliance. Of the 3 women with emerging symptoms, 1 had predominantly anxiety and two had emerging hypomania. Three of those receiving placebo completed the 52 weeks (75%), with 1 (25%) discontinuing due to irritability. While there was no statistically significant difference between groups, the numbers in cells are admittedly small. Three women conceived and ultimately delivered healthy babies (1 in the DHA group, 2 in the placebo group).

Additionally, psychotherapy is a potential alternative treatment during pregnancy or postpartum for women with only mild-to-modcrate depressive symptoms. These are best managed with cognitive behavioral therapy, interpersonal therapy, or family focused therapy.[81-85]

While electroconvulsive therapy (ECT) is another nonpharmacological option for treatment during pregnancy, it should be reserved for severe cases, such as women hospitalized in a severe vegetative state or psychotic episode. As noted above, there is no indication of teratogenesis associated with ECT,[86] and the treatment is considered relatively safe during pregnancy.[87]

Women and their providers should carefully discuss treatment options during pregnancy and the postpartum period. While most experts agree that medication use during the first trimester should be minimized, there are options for conservative treatment in the remainder of pregnancy and postpartum. Additionally, there are nonpharmacological options such as omega-3 fatty acids or psychotherapy, and ECT for severe cases. Women and their partners should plan ahead to ensure adequate social support and assistance with infant care to minimize sleep disruption and stress that may increase the risk for relapse.

Summary

While the prevalence of bipolar disorder is comparable in men and women, there are several aspects of the disorder that require unique consideration in women. Women with bipolar disorder should be checked for hypothyroidism and, if prescribed antidepressants, carefully watched for rapid cycling or a mood switch to mania, hypomania, or a mixed state. Several medications interact with oral contraceptives or increase the risk of developing PCOS, and this should be considered when choosing a medication for women of childbearing potential. Consideration of possible pregnancy is essential, and should be planned in advance whenever possible. Decisions about treatment while trying to conceive and once pregnant, risks of medication to the fetus at various points of development must. be balanced against the risks of not. treating, which is also detrimental to both fetus and mother. However, medications differ in their teratogenic risk, and there are options for discussion. The postpartum period is a time of especially high risk, and most, women should resume pharmacotherapy. The decision to breastfeed must also consider the adverse impact of sleep deprivation in triggering mood episodes. 'The care of women with bipolar disorder r requires a strong clinician-patient relationship and collab1 orativc planning and decision-making.

Table I

Risks of common psychotropics for the treatment of bipolar disorder during pregnancy and while breastfeeding.

Médication	Risk In pregnacy	Breastfeeding risk
Lithium	Risk for cardiovascular abnormalities, most notably Estein's anomaly, incidence ranges from 4 %-12 %. Risk for toxicity.	Contraindicated for use with breastfeeding by the American Academy of Pedidtrics Cummitee on Drugs. High concentrations of the drug present in infant serum and breastmilk (24%-72% of maternal serum concentration). Can be associated with rapid dehydration in febrile infant. Other detrimental effects can include lethargy, hypothermia. hypotonia, and T-wave modifications on electrocardiogram.
	Other risks include premature delivery: floppy infant syndrome, transient neurodevelopmental deficit, nephrogenic diabetes insipdus, thyroid dysfunctions, and rarely polyhydramnios.
Valproate (VPA)	VPA use in pregnancy associated with an approximately 2-5 fold increased risk of major malformations, especially if administered in the first trimester. Overali incidence of major malformations is 11%. Most common risks include spina bifida (1%-5% risk), developmental retardation, skeletal malformations, and cardiovascular abnormalities. “Fetal vaiproate syndrome” includes cardiovascular, craniofacial, urogenital, digital, repiratory tract abnormalities and developmentaI delay. lncreased risk for miscarriage andfor stillbirth.	CIassified as compatible with breastfeeding by the American Academy of Pediatrics Committee on Drugs.
		Present inthe serum of breastfed infants (%<1-10%). In one 3-month infant, thrombocytopenic purpura and anemia were attributed to valproate exposure through placenta and breast milk.
Carbamazepine (CBZ)	Overall incidence of CBZ-related fetal malformations is 5.7%. They include microcephaly, other cranio-facial skeletal deficits, growth retardation, and cardiac deficits. The association between neural tube defects, cardiovascular and urinary tract anomalies, and cleft palate in neonates exposed to the compound is called “carbamazepine syndrome”. Increased risk for spina bifida (0.5%-1.0%). Also assooated with fatal vitamin-K deprivation, resulting in increased risk for neonatal bleeding and midfacial abnormalities.	Classified as compatible with breastfeeding by the American Academy et Pediatrics Committee on Drugs. Present in breast milk (7%-95%) and serum (6%-65%) of breastfed infants; no widespread adverse affects noted for breastfed infants. Carbamazepine is metabelized rapidly in newborns.
Lamotrigine (LTG)	Lesser risk for malformatiorn compared with other antkonvulsants. In epilepsy, birth defects observed in 29% of pregnant women exposed to LTG monotherapy. One report that doses >200 mg/day may be associated with increased risk for major malformations.	Present in the serum of breastfed infants (range from 23%-33% of maternal leveIs); monotoring for rash in infants recommended. The manufacturer does not recommend use during breastfeeding.
First generation antipsychotics	Important treatment options for acute mania during pregnancy; Iow doses for ongoing therapy for those who want a ess teratogenic option.	No adverse effects noted in breastfed infants exposed to these drugs.
	Association with hyperprolactinemia.
Secon-generation antipsychotics	Little safety data on use in pregnant women. Preliminanj indicaVon is that the atypical antipsychotics are net associated with increased rates of major structural malformations. These drugs may impact fetal development because of known association with metabolic abnormalities in the mother; olanzapine associateci with gestational diabetes and preeclampsia.	These drugs are secreted into breast milk in relatively small amounts.
Antidepressants	Labeling added in 2004 for physicians to consider tapering use of these medications near the end of the third trimester. This action aims to reduce jitteriness, irritability, feeding difficulties, trouble breathing, and other problems sometimes seen in newborns whose mothers took selective serotonin reuptake inhibitors (SSRIs) or serotenin and norepinephrine reuptake inhibiters shortly before delivery. Such problems usually are mild and disappear by 2 weeks of age. Paroxetine asseciated with increased risk cf cardievascular malformations, espedally ventricular septal defects. Rare pulmenary hypertension with SSRIs, especially affer 20 weeks exposure.