Management of Depression and Psychoses in Pregnancy and in the Puerperium

Mood Disorders in Pregnant and Postpartum Women

Mood disorders are common in pregnant and postpartum women, although prevalence rates vary among studies. Women with unipolar major depressive disorder (MDD) have MDEs and women with bipolar disorder (manic-depressive illness) may have both MDEs and episodes of mania at different times. The Agency for Healthcare Research and Quality found a best estimate for risk for an MDE at any point in pregnancy to be 12.7% (95% confidence interval [CI], 7.1–20.4); the best estimate for risk during the 3 months after delivery was 7.1% (95% CI, 4.1–11.7). Because of the overlap between CIs, these data do not necessarily support a difference between pregnant and postpartum women in their risks for unipolar MDD. The course of illness can vary: some women become depressed in pregnancy and continue to be symptomatic into the postpartum period; others improve shortly after delivery. About half of women who are depressed in the postpartum period had an onset of illness after the delivery.

The typical age of onset for bipolar disorder is in the late teens to early 20s. Bipolar disorder occurs in about 1% to 2% of the population; its prevalence is similar in pregnant and nonpregnant women. The risk of a mood episode for a woman with bipolar disorder is clearly increased during the postpartum period in comparison to the risk for nonpregnant ^, and pregnant ^, women with unipolar MDD.

Major Psychotic Disorders in Pregnant and Postpartum Women

Women with schizophrenia and schizoaffective disorder have chronic conditions with symptoms that, ideally, are controlled by antipsychotic medication and mood-stabilizing drugs. Vulnerability to either schizophrenia or schizoaffective disorder is not affected by pregnancy or being in the postpartum period, although symptoms may be under slightly better control in pregnancy. However, women who become pregnant are vulnerable to exacerbation of symptoms if medications are stopped. The stress associated with either pregnancy or being in the postpartum period can also exacerbate symptoms and feed psychotic delusions.

Major Depressive Disorder

As with many psychiatric disorders, biologic (e.g., genetic) vulnerability, stress, and trauma are some of the factors that underlie the risk for MDD. There are multiple theories of depression, and they are not necessarily exclusive. For example, the presence and activity of neurotrophic and growth factors are enhanced by antidepressant treatment, which suggests dysregulation of these systems in depression. People with depression show associations between negative information and activation of brain areas such as the anterior cingulate cortex, the amygdala, and the hippocampus and lower activity in regions related to reward, such as the ventral striatum. A wealth of literature also associates mood symptoms with elevated levels of cortisol and corticotropin-releasing hormone (CRH). Hypersecretion of these hormones can lead to reductions in the size of critical brain regions such as the hippocampus. Importantly, serotonin is involved in regulation of the hypothalamic-pituitary-adrenal (HPA) axis. Hyperactivity of the HPA axis may increase the risk of neuroinflammation and lead to microglia activation, which is also associated with MDD. ^,

Psychosocial and familial factors and medical history are related to the risk for depression after delivery. In particular, a personal history of depression, a family history of depression, and the lack of partner support increase the risk for postpartum depression. Specific triggers or biologic deficits that increase the risk for mood symptoms and mood disorders in pregnancy or the postpartum period are not well delineated but are areas of active investigation. Some research suggests dysregulation in the HPA axis, as found for MDEs occurring outside the perinatal period. ^, Women who experience postpartum mood symptoms, as compared with postpartum women who do not, may be more likely to have elevated CRH levels in pregnancy, although not all studies find this. The elevations of placental CRH in pregnancy lead to a refractory period of HPA reactivity after delivery, although women who are postpartum and depressed have an extended, blunted response. In addition, the activity of CRH neurons is mainly regulated by γ-aminobutyric acid (GABA) inhibition via GABA type A (GABA _A ) receptors. Neuroactive steroids, such as progesterone and its metabolite, allopregnanolone, are implicated in the pathophysiology of postpartum depression. Allopregnanolone is a positive allosteric modulator of GABA _A receptors. GABA signaling plays an important role in postpartum depression in various animal ^, and human studies, and low levels of allopregnanolone are associated with development of postpartum depression.

It is also notable that the elevation in adrenocorticotropic hormone that occurs after delivery coincides with the development of mood symptoms. Uncoupling between adrenocorticotropic hormone and the cortisol response is found among women depressed after delivery. Other studies find that women who are depressed in pregnancy have mild elevations of thyroid-stimulating hormone and lower thyroxine levels, indicative of mild hypothyroidism. Lower levels of triiodothyronine and thyroxine in pregnancy are associated with postpartum depressive symptoms.

Researchers have probed changes in neurocircuitry related to an MDE in the postpartum period. Although there are differences among studies and within regions, there appears to be diminished cortico-cortic and cortico-limbic connectivity in women with postpartum depression compared to those without symptoms. After provocation with an infant cue, such as a cry, the amygdala response, particularly on the right, appears blunted in some studies.

Bipolar Disorder

The biologic underpinnings of bipolar disorder are being actively explored, particularly through the use of brain imaging and genetics. Although not definitive, differences in the amygdala, paralimbic structures, and frontal cortices, as well as in the connections between these structures, are found between individuals with and without bipolar disorder. ^, The amygdala appears to be hyperactive in patients with mania, while “top-down” control from the prefrontal cortex is reduced. Dysfunction in the prefrontal cortex may be a trait deficit, because it is also found in depressed and euthymic mood states of people with bipolar disorder. Genetics probably plays a role, because selective nerve growth factors are associated with amygdala size. ^, Treatment with agents such as lithium may promote cellular growth in certain regions, perhaps explaining the long-term therapeutic effects of these agents.

The state of pregnancy, per se, does not increase the risk for relapse among women with bipolar disorder. ^, Rather, discontinuation of a number of medications used to treat bipolar disorder, many of which are teratogenic, is a powerful cause for illness relapse in pregnancy. ^, On the other hand, numerous studies found that women who suffer from bipolar disorder, even if they are treated, are at high risk for relapse during the postpartum period compared with nonpregnant or pregnant women with bipolar disorder. ^, The high dosages of gonadal steroids expressed by the placenta may contribute to mood stabilization in pregnancy and deterioration in mood after postpartum withdrawal. Some hypothesize that dysregulation of sleep patterns contributes to postpartum relapse, but evidence in support of this is difficult to find. ^, Of course, mania and depression are characterized by sleep difficulties, so there is some circularity in this hypothesis, and careful longitudinal studies are needed.

Schizophrenia and Schizoaffective Disorder

As with the disorders already discussed, definitive causes of schizophrenia and schizoaffective disorder, which are major psychotic disorders, are not known, although genetic vulnerability, epigenetic factors, and, to a lesser extent, environmental stressors contribute to illness expression. For decades, the “dopamine hypothesis” dominated explanations for psychotic disorders such as schizophrenia and schizoaffective disorder. This model assumed that hyperactivity of the dopaminergic system was at the root of psychosis. The model was largely based on the antidopaminergic properties of first-generation antipsychotic agents, but it was also a result of the observation that psychotic patients are sensitive to dopamine agonists when administered in provocation studies. Over the years, the dopamine hypothesis has been modified to consider that a hyperresponsive dopamine system may be at fault, suggesting roles of glutamate and GABA and their receptors and circuits. For example, GABA is a major inhibitory neurotransmitter and inhibits release of dopamine. Loss of GABAergic inhibition, either because of hypoactivity of glutaminergic receptors or neurons in critical brain regions or because of destruction of GABAergic neurons as a result of environmental insult or infection, releases the tonic inhibition of dopamine and hence increases dopamine activity.

A frequently cited clinical finding among women who suffer from schizophrenia or schizoaffective disorder is that their illness improves during pregnancy and then worsens after delivery. Some also claim that women with psychotic disorders, particularly when pregnant, require lower dosages of antipsychotic medications. These findings have been attributed to the antidopaminergic effects of estrogen, which leads to the dampening of dopamine activity. The situation is reversed after delivery, when estrogen levels diminish and higher dopamine activity ensues.

Major Depressive Episodes and Major Depressive Disorder

A diagnosis of MDD is made when a woman has episodes of major depression but does not have episodes of mania or hypomania. An MDE includes symptoms of low mood and diminished capacity to experience pleasure (see Table 67.1 ). There are nine candidate symptoms for MDE (oversleeping or undersleeping are one item, and agitation or mental slowing are one item), although a minimum of five, including either low mood or diminished pleasure, are required for diagnosis. Symptoms should be present for most of the day and for most days during a 2-week interval. Women who are depressed frequently cite problems with energy, they may either oversleep or sleep too little, and they overeat or have a poor appetite. Women who are pregnant and not depressed also experience changes in energy, sleep, and appetite. Thus a diagnosis of MDE in pregnancy requires that a clinician ask specifically about mood and other nonbehavioral symptoms. Knowledge about mood symptoms in conjunction with behavioral symptoms will allow the clinician to render an accurate diagnosis of MDE.

Depression screening can be helpful in identifying women who are symptomatic. The screening process alone may have a positive effect on depressive symptoms in pregnant and postpartum women. Greater benefits of screening are gained if practitioners have the skills and resources to deliver adequate interventions or to provide appropriate referrals. Well-validated screening instruments measure general distress and dysphoria and are often less specific for an MDE. Appropriate tools for assessment of depressive symptoms in pregnant women include the Edinburgh Postnatal Depression Scale (EPDS), the Patient Health Questionnaire-9, the Inventory of Depressive Symptomatology (IDS), or the Primary Care Evaluation of Mental Disorders (PRIME-MD) Patient Health Questionnaire. The EPDS is designed specifically for use with postpartum women, although it is less comprehensive than the IDS and does not include all of the criteria for MDD. The EPDS has 10 items and a mood and anxiety subscale; scores range from 0 to 30. The IDS measures severity of cognitive features of depression and anxiety and atypical depressive symptoms, such as overeating and oversleeping (rather than undersleeping and undereating); it has self-report and clinician-administered versions, as well as a brief version. The PRIME-MD records diagnoses of MDE and minor depressive disorder and has been used in obstetrics-gynecology settings; it takes 5 to 20 minutes to complete.

Bipolar Disorder

A depressive episode is also experienced by individuals with bipolar disorder. However, for a diagnosis of bipolar disorder, a woman must have experienced an episode of mania or hypomania (see Table 67.1 ) at some point in her life. Thus the defining feature for bipolar disorder is at least one episode of mania or hypomania, with or without a history or a current MDE. It is not unusual for a person to present initially in an MDE and then experience an episode of mania later in the course of her illness. This occurs more often in women than in men and reflects the fact that women with bipolar disorder are more likely to have MDEs and depressive symptoms relative to manic episodes during the course of their illness. Psychotic symptoms, such as auditory hallucinations or delusions, can also occur in women who have unipolar MDD or bipolar disorder, schizoaffective disorder, or schizophrenia. However, psychotic symptoms are not the central feature in either MDD or bipolar disorder, as they are among individuals with schizophrenia or schizoaffective disorder.

Many women with bipolar disorder were diagnosed before presenting in pregnancy and are usually able to provide this information and their lifetime treatment history to their obstetrics provider. However, the first onset of bipolar disorder may occur after delivery or even during late pregnancy. Often these women are described as having “postpartum psychosis.” Women with bipolar disorder constitute the largest diagnostic group among women with postpartum psychosis. In fact, women who develop psychiatric symptoms rapidly, even if the symptoms are those of depression, are highly likely to manifest bipolar disorder in the upcoming months and years. Whereas a postpartum woman may be exhausted, a woman with mania has excessive energy and is unable to sleep even if she is given the opportunity to rest. Other symptoms of mania are listed in Table 67.1 .

Screening for bipolar disorder is complicated, because an individual may present either in a manic or in a depressed state. Only two screening questionnaires have been tested in pregnant and postpartum women, and the evaluation has been limited to mania. The Highs self-rating scale screens for an elevated mood that, depending on severity, may be indicative of mania. It was specifically developed for use in postpartum women. The Mood Disorder Questionnaire was developed for use in primary care and was evaluated in one perinatal study. Both instruments are short and easy to administer. A woman who screens positive or develops the symptoms in Table 67.1 should be immediately evaluated by a psychiatrist.

Schizophrenia and Schizoaffective Disorder

Women who experience psychosis and are not in an episode of mania or depression may have a chronic psychotic condition such as schizophrenia or schizoaffective disorder. Schizoaffective disorder is differentiated from schizophrenia in that the former condition requires mood symptoms, including episodes of mania or major depression, that are prominent, although psychosis may also occur without mood symptoms. Mood symptoms can occur in women with schizophrenia, but they are not a cardinal feature of that disorder and are not as prominent. Psychosis is chronic and central to both of these diagnoses.

The peak period for the onset of schizophrenia or schizoaffective disorder is the late teens and early 20s. Thus many women who are pregnant and suffer from schizophrenia or schizoaffective disorder can provide diagnostic and treatment history. Although there are no screening scales for schizophrenia in pregnant or postpartum women, the development of delusions or hallucinations should trigger a psychiatric evaluation. Because many individuals and their doctors are uncomfortable with the use of psychotropic medication in pregnancy, women may stop pharmacologic treatment, which can precipitate an illness relapse. Postpartum women with schizophrenia or schizoaffective disorder may be at greater risk for relapse during the puerperium because of biologic factors, such as the withdrawal of reproductive hormones at delivery.

Unipolar Major Depressive Disorder

Limited research suggests that high levels of stress are associated with spontaneous pregnancy loss in chromosomally normal offspring. Because an MDE is a stress-related illness, similar results may be expected for women with MDD. Unfortunately, studies looking specifically at depression are small and the results are mixed, making it difficult to draw definitive conclusions.

Much of the research addressing the relationship between depression and pregnancy outcomes assessed women with depressive symptoms rather than a diagnosed depressive disorder. This distinction is important, because people with nondepressive psychiatric conditions can also have elevated scores on depression screening measures. A meta-analysis of 23 studies (all but one were prospective observational studies) found that untreated depression was associated with preterm birth (odds ratio [OR] = 1.56; 95% CI, 1.25–1.94) and low birth weight (OR = 1.96; 95% CI, 1.24–3.10). Importantly, preterm birth rates were not statistically different when comparing studies based on a clinical diagnosis of depression versus depressive symptoms from a self-administered questionnaire. Also important is the inability of the authors to control for critical exposures such as licit and illicit substances that are associated with preterm birth and low-birth-weight deliveries. A cohort study ( n = 7267) considered “depressive symptoms” as the exposure and used the EPDS to screen pregnant women and found that mildly elevated scores (cutoff >10) but not higher scores (>12) were associated with preterm birth, very preterm birth, small for gestational age (SGA) fetus, and low-birth-weight delivery. Other studies that included a depression diagnosis, rather than depressive symptoms, did not find MDE to be an independent and robust risk factor for an adverse birth outcome. ^, Though current evidence does not conclusively support MDE in pregnancy as a significant risk factor for preterm birth or for delivery of a low-birth-weight infant, women with elevated scores on a distress measure such as the EPDS appear at risk for such events. In any event, women who report symptoms merit further evaluation for personal risk as well as perinatal risks.

Bipolar Disorder

Most research on birth outcomes and other postnatal complications derive largely from linked administrative databases that allow investigators to obtain sufficient sample sizes for relatively uncommon disorders (∼1% for bipolar) and birth outcomes. Such studies have important limitations. For example, most are not able to control adequately for smoking and illicit drug use. The possible role of pharmacologic treatments in poor birth outcomes has been inadequately addressed because of heterogeneity in medication use and limited cohort sizes. With these caveats in mind, studies find that, compared to gravidas without bipolar disorder, those with the condition are at higher risk of adverse birth outcomes, including preterm birth, ^, microcephaly, gestational hypertension, and antepartum hemorrhage.