Women's Primary Health Grand Rounds at the University of Michigan; Timothy R.B. Johnson, MD, and Barbara Apgar, MD, Series Editors

Postpartum Depression

Historically, the connection between childbirth and psychiatric illness has been well-recognized. In 460 BC, Hippocrates described "puerperal fever," theorizing that suppressed lochial discharge was transported to the brain, where it produced "agitation, delirium and attacks of mania."¹ The 11th century writings of the gynecologist Trotula of Salerno speculated: "if the womb is too moist, the brain is filled with water, and the moisture running over to the eyes, compels them to involuntarily shed tears."² Attempts to describe and classify postpartum mental illness became more systematic in the mid-19th century, when Esquirol wrote of the "mental alienation of those recently confined and of nursing women."² Additionally, accounts of puerperal psychosis and depression are specifically delineated by Marce in his 18th century Treatise on Insanity in Pregnant and Lactating Women.²

Nonetheless, while more common than gestational diabetes, preeclampsia, and preterm delivery, postpartum depression has received much less attention in contemporary medical literature, training, and clinical practice. Although both the academic and lay press have recently increased the focus on postpartum depression, this condition remains a frequently overlooked illness despite its potentially devastating consequences. Debate continues about its cause, definition, diagnostic criteria, and even its existence as a distinct entity.

INCIDENCE

Depression is an exceedingly common disorder, affecting 15% to 25% of the population^2,3 and representing a yearly economic burden of $44 billion.⁴ Overall, depression is frequently undetected; indeed, fewer than 25% of patients suffering from mental illness are actually under the care of a mental health specialist.^2,4 Depression is twice as common in women as it is in men, with its peak incidence during the primary reproductive years (25 to 45). The link between reproductive status and depressive illness is further evidenced by the high frequency of depression during the premenstrual phase, the perimenopausal period, and the immediate postpartum period.⁴

As one of the major physical, psychological, and social stresses of a woman's life, childbirth is gaining growing recognition as a major risk factor in the development of mental illness. Early descriptive studies have been bolstered by a series of reports that document the temporal relationship between psychiatric admissions and childbirth.^1,5,6 Using data from more than 35,000 deliveries with sequential 90-day intervals over a 2-year period preceding and following the delivery, Kendell and colleagues clearly demonstrate a sevenfold increase in the risk of psychiatric hospitalization in the first 3 months after delivery.5 In this study, the risk of psychosis was 22 times higher than the prepregnancy rate, and the calculated relative risk (RR) of childbirth for the development of psychosis was 16. However, 87% of the admissions were for affective disorders, with the majority of diagnoses being major depression.5,6

This significant incidence of depression--and the remarkable peak in the number of new cases diagnosed shortly after delivery--makes it clear that the puerperal period is unique in the development of mental illness. It is estimated that of the approximately 4 million births occurring annually in this country, 40% are complicated by some form of a postpartum mood disorder.⁷

PRESENTATION

Although the current literature divides the spectrum of postpartum mood disorders into three distinct categories, these classifications frequently blend at the margins. At the mildest end of the spectrum is the "maternity blues" or "baby blues." Because this condition arises after 40% to 85% of deliveries, practitioners and patients often view it as a "normal" phenomenon. Nonetheless, patients and their families are distressed by the patients' depressed mood, irritability, anxiety, confusion, crying spells, mood lability, and disturbances in sleep and appetite. These symptoms peak between postpartum days 3 and 5, and typically resolve spontaneously within 24 to 72 hours. The primary treatment is supportive care and reassurance about the transient nature of the condition.⁸

At the other end of the spectrum is the truly devastating puerperal psychosis. A comparatively rare disease, it complicates only 0.1% to 0.2% of deliveries--but this is 12 to 14.5 times the prenatal incidence of psychosis.^6,9 Symptoms generally present within the first 4 weeks postpartum, when the risk of hospitalization is 22 times greater, but can manifest up to 90 days after delivery. A second, smaller, peak in incidence is evident at 18 to 24 months. Patients suffering from puerperal psychosis are severely impaired, suffering from hallucinations and delusions that frequently focus on the infant dying or being divine or demonic. These hallucinations often command the patient to hurt herself or others, placing these mothers at the highest risk for committing infanticide and/or suicide. Most of these patients suffer from affective disorders (primarily bipolar illness), but schizophrenia and even organic brain syndromes are also diagnosed.⁷

Between these two extremes is postpartum depression, which is increasingly recognized as a unique and serious complication of childbirth (Table 1). Its insidious onset and chronic course complicates 10% to 15% of all deliveries^7,10,11 and a staggering 26% to 32% of all adolescent deliveries.¹² The majority of patients suffer from this illness for more than 6 months and, if untreated, 25% of patients are still depressed a year later.¹³ Postpartum onset is the index depression episode in more than 50% of cases, again underscoring the unique triggering effect of childbirth. More than 60% of patients have an onset of symptoms within the first 6 weeks postpartum,¹⁴ providing the primary care physician with the perfect opportunity for diagnosis.

As these patients often suffer from delusions and suicidal tendencies, the consequences of this disease to both mother and child are significant. Furthermore, depressed mothers have an increased risk of relapsing and/or continued psychiatric illness. Both Kumar and Robson¹⁰ and Philips and O'Hara¹⁵ found that during a 4-year follow-up period, approximately 80% of patients sought help again for psychiatric complaints. In addition, depressed mothers often show a more negative attitude toward their children, and a debilitated new mother puts significant emotional and perhaps economic burdens on family relationships. The patients themselves are often the most sensitive to these consequences. A recent study reports that 32% of women who suffered an episode of postpartum depression dramatically changed their future childbearing plans, re-sorting to adoption, abortion, or in some cases even sterilization.¹⁶

The children of these women also suffer because of this illness. Hospitalization of a mother in the early postpartum period increases the incidence of emotional disturbance in children. Conflicting results have been published regarding the effects on the social and emotional development of the children of patients who are not hospitalized, but long-term follow-up studies of up to 4 years suggest that depressive episodes in a mother during the postpartum period were linked to poorer cognitive test scores in their children.¹¹

One of the primary risks of postpartum depression is continued or relapsing illness. Practitioners must recognize that the highest risk of relapse is in subsequent deliveries, where the recurrence risk is 1:3 to 1:4.⁸ The risk of recurrence may also be correlated to the severity of the initial symptoms; in a subset of women with onset of psychotic symptoms within the first 24 months postpartum, the recurrence risk approaches 100%.¹⁷

DIAGNOSIS

Given this incidence and impact, identification of patients suffering from postpartum depression should be a priority for all physicians who treat women. The diagnostic criteria for a major depressive disorder are no different in the postpartum period, with the exception that symptoms must be present for more than 2 weeks postpartum to distinguish them from the "baby blues." Diagnosis requires that the patient experiences either dysphoric mood or anhedonia most of the day, nearly every day, for at least 2 weeks. Additionally, at least four of the following symptoms must be present: difficulty concentrating or making decisions; psychomotor agitation or retardation; fatigue; changes in appetite and/or sleep; recurrent thoughts of death or suicide; feelings of worthlessness or guilt, especially focusing on failure at motherhood; excessive anxiety, frequently focusing on the child's health.^2,18

The predominance of anxiety and delusions among these patients often leads authors to describe postpartum depression as "atypical."⁸ Weight and appetite changes in recently delivered (and often breast-feeding) women are expected, and sleep deprivation is universal in early motherhood. Therefore, the detection of pathologic changes requires specifically directed questions. For example, the physician might ask the mother if she is tired and able to sleep when the infant sleeps, or if food appeals and tastes good to her.

As with other common complications of pregnancy, physicians must remember that all women are at risk. However, certain factors have been identified that may place a patient at particular risk. There is a clear association between postpartum depression and a family history of depression--especially a prior personal episode of depression--that may raise the patient's risk as high as 30%. One episode of postpartum depression may result in a recurrence risk of up to 70%.⁴ Women experiencing a poor marital relationship (particularly an abusive relationship), a lack of other social supports, and/or child care stressors are also at increased risk. The comorbidities of substance abuse and anxiety or somatization disorders substantially increase the risk of postpartum depression. Primiparous women have been reported to suffer from postpartum depression more frequently, but other studies refute this; the documented finding that patients frequently change their reproductive plans after suffering from postpartum depression confounds the research. Other controversial risks reported in the literature include breast-feeding and obstetric factors such as the length and difficulty of labor, multiple gestation, and advanced maternal age.^14,19,20 Postpartum depression is a cross-cultural phenomenon, and likewise has not been associated with socioeconomic class or education level.^7,14

EMOTIONAL/HORMONAL CONSIDERATIONS

Multiple investigations into the etiology of postpartum depression have not reached a consensus. The aforementioned risk factors increase stress, and have been postulated as psychological factors leading to depression. Some psychologists have also ascribed the etiology of postpartum depression as inherent to the patient's psychological construct. The cognitive model postulates that a patient's negative view of the world and herself leads to depression through low self-esteem and disturbed relationships, and that patients with abnormal attitudes of self-control may develop a learned helplessness leading to depression.⁷ Nonetheless, all patients with a negative outlook on the world do not become pathologically depressed, and there is no well-defined etiology for depression in general. Moreover, there is no clear evidence that the mechanisms that produce nonpuerperal depression operate in the postpartum period.

The two leading biologic theories for the cause of nonpuerperal major depression lie in the deregulation of the neurotransmitters serotonin and the other biogenic amines such as norepinephrine, epinephrine, and dopamine. The role of these factors has also been investigated in postpartum depression. Tryptophan is the main precursor for serotonin, and it has been postulated that low circulating tryptophan levels lead to low serotonin levels, which may lead to depression. Despite three separate studies demonstrating decreased tryptophan levels in women suffering from postpartum depression, a double-blind, placebo-controlled trial replacing tryptophan showed no effect.⁷ Catecholamines have not been studied as thoroughly, but low norepinephrine levels and the severity of postpartum depression have been correlated.⁷

Other biologic factors examined in the etiology of postpartum depression have included hormonal factors known to change in the puerperium. Postpartum thyroiditis and hypothyroidism are not uncommon phenomena. Patients with documented hypothyroidism do have a mild increase in depressive symptoms, but studies of patients with postpartum depression have demonstrated conflicting findings regarding thyroid function.⁷ The hypothalamic pituitary axis is also implicated, and has been widely investigated in depression. Cortisol levels rise in pregnancy, peak in labor, then fall dramatically following delivery, after which they decline more slowly to normal levels.²¹ A strong correlation between cortisol levels and the incidence of depression is not apparent; both high and low levels of cortisol have been documented during the postpartum period in women suffering from postpartum depression.^7,14 Prolactin levels, which also rise throughout pregnancy and fall more slowly after delivery, have also been considered as a potential factor in postpartum depression. Once again, conflicting results have been obtained; higher levels of prolactin have been found in women with depressed mood on postpartum days 2, 4, and 6, while lower levels have been correlated with a depressed mood in postpartum week 6.⁷

Unique to the puerperium are the rapid changes in the levels of ovarian steroids. These hormones fall from very high levels at term to near follicular levels within 48 hours. The rapid withdrawal of progesterone--whose metabolite hydroxy-5a-dihydropro- gesterone is a potent, barbiturate-like ligand of g-aminobutyric acid (GABA) receptors22--occurs with the delivery of the placenta, and has been postulated as a causative factor. The magnitude of decline of progesterone from 38 weeks' gestation to 1 day postpartum was associated with depressed mood in the first week, while a later study showed no association between progesterone levels and postpartum depressive symptoms. An interesting finding was lower levels of progesterone in breast-feeding, depressed women versus nondepressed women, while a reverse association was found in nondepressed women.⁷ These findings suggest some interplay between progesterone and prolactin, and a need for further research into this connection.

Estrogen levels, like progesterone, fall by 90% to 95% in the first 48 hours following delivery. There is recent accumulating evidence of estrogen's important role in memory and cognition, as well as mood. Moreover, there is a well-established body of literature regarding estrogen's profound effects on brain function. These effects are as basic as causing new neuronal spine and dendrite formation, to very specific interactions on all of the relevant neurochemical systems.²³ The specific effects are best characterized in the dopamine system, where estrogen increases dopamine turnover through the regulation of tyrosine hydroxylase (the rate limiting synthetic enzyme), degradative enzymes such as monoamine oxidase, and turnover of dopaminergic receptors. Estrogen has similar effects on the serotonergic and gabaminergic systems, increasing the sensitivity of neurons to norepinephrine and inducing the messenger ribonucleic acid for endogenous opiates such as proenkephalin.²³ Despite this circumstantial evidence, however, no link between estrogen levels and postpartum depression has been established. Indeed, both low predelivery levels and high postpartum levels of estrogen have been reported in depressed women.⁷

Taken together, the data suggest that the etiology of postpartum depression is multifactorial. Causative components include both organic and conditional changes arising from both parturition as well as the mother's surrounding situation, which likely combine to influence the patient's psychological function.

Despite the possibility of a unique etiology for postpartum depression and the lack of controlled comparative therapeutic trials, there is no reason to believe it responds differently to treatment than other types of depression. As in other episodes of depression, early identification and treatment are the keys to successful therapy. Treatment of depression involves three phases--acute treatment (6 to 12 weeks) aimed at remission of symptoms, continuation treatment (4 to 9 months) aimed at stabilization and recovery, and maintenance treatment aimed at preventing recurrence in patients with prior episodes.¹⁸

Postpartum depression is successfully treated with medications, psychotherapy, or a combination of both.^7,14 Pharmacologic treatment is preferred to psychotherapeutic intervention in patients with more severe or chronic symptoms, prior episodes or family histories, or a prior response to treatment. Medications have the advantage of being less costly and time-consuming. Psychotherapy should be added in patients with more severe depression, chronic psychosocial problems, incomplete response to medication, or evidence of a concurrent personality disorder.¹⁸

Criteria for consultation or referral include clinician preference, suicidal or homicidal ideation, presence of psychotic symptoms, severely impaired functioning (including obsessing about or avoiding the infant), failure to respond to an antidepressant treatment trial, and comorbid substance abuse.¹⁸ Primary care physicians who initiate treatment of patients' postpartum depression should be familiar with the dosages and side effects of one or two drugs from the major categories listed in Table 2. In general, the tricyclic antidepressants--which block norepinephrine and to a lesser degree serotonin reuptake--are effective, but treatment is burdened with side effects such as weight gain, sedation, dry mouth, and cardiovascular effects. These may be particularly unacceptable in the postpartum period. Moreover, these are very toxic medications, making them particularly dangerous in patients with high overdose potential. The selective serotonin reuptake inhibitors (SSRIs) have fewer and less prominent side effects and provide a wider margin of safety. Monoamine oxidase inhibitors are rarely used now owing to the obligatory dietary restrictions, while newer drugs such as bupropion and venlafaxine have not been tested in the postpartum period but only under psychiatric care.

Patients need frequent monitoring of side effects and treatment response, with a formal 6-week evaluation of partial or nonresponders; this should involve a reevaluation of the diagnosis, compliance with medication, and the need to increase dosing or to change treatment. Electroconvulsive therapy is an important, effective therapeutic intervention that can be useful in treating recalcitrant cases of postpartum depression.

As with all postpartum treatment decisions, special consideration must be given to breast-feeding women. The tricyclic antidepressants have been extensively studied, and no negative effects or detectable serum levels have been demonstrated in the infants of treated mothers. The newer SSRIs have not yet been fully evaluated. Their long-half life is of special concern, as neonates have only 33% to 50% of the adult hepatic metabolic capacity. In fact, treatment of a mother with fluoxetine has been associated with both colic and high serum levels in the infant.²⁴ Therefore, the current recommendations are to treat breast-feeding depressed patients with tricyclics at the minimum effective dose, and to evaluate the infant's level if it is less than 10 weeks old or if symptoms develop.²⁴

To date, prophylactic treatment for women with perceived risks of postpartum depression has not been widely examined. Several uncontrolled studies have attempted to prevent recurrent severe postpartum depression with antidepressants,²⁵ progesterone,²⁶ and estrogen¹⁷ with promising results. However, the only prospective treatment trial aimed at a potential etiologic factor in postpartum depression was a double-blind, placebo-controlled trial by Gregoire and colleagues,¹³ who randomly treated 61 patients affected by postpartum depression with either transdermal estrogen or placebo for 6 months. They showed a significant, rapid, and stable fall in the patient- and clinician-rated depression scores of the estrogen-treated group, with only 50% and 20% of treated patients depressed at 1 and 4 months, respectively. By contrast, 74% and 69% remained depressed in the placebo group. This study represents an important development in the understanding and treatment of postpartum depression. While it needs replication prior to widespread clinical use, this study raises important issues about the general antidepressant nature of estrogen.

Practitioners have the ability to decrease the impact and devastation of postpartum depression by following some simple guidelines for its prevention and treatment. Information about the incidence and the warning signs of postpartum depression should be an intrinsic part of prenatal education. This ideally should include information about mothering classes that may help patients' expectations and suggest ways to make use of existing support systems. Even more importantly, clinicians need to identify patients who have suffered prior episodes of depression, have poor support, or who have other comorbidities putting them at highest risk for postpartum depression. These patients need careful postpartum follow-up, and consideration should be given to prophylactic treatment of those at highest risk.

During the postpartum phase of care, clinicians need to recognize the symptoms of depression and to realize that patients are embarrassed about feeling unhappy during a time when society expects them to be elated. Therefore, it is important to ask patients specifically about their mood and adjustment. The Edinburgh Postnatal Depression Scale is an excellent tool for use in patient evaluation.

CONCLUSION

Postpartum depression is a common, frequently unrecognized, yet devastating disorder. The keys to successful treatment are early identification and intervention, both supportive and pharmacologic. These treatments are effective, and the ability to lessen the impact of this disease is congruent with the primary care provider's role.

REFERENCES

1. Thurtle V. Post-natal depression: the relevance of sociological approaches. J Adv Nurs. 1995;22(3):416-424.
2. Steiner M. Postpartum psychiatric disorders. Can J Psychiatry. 1990;35:89-95.
3. Yonkers KA, Gullion C, Williams A, et al. Paroxetine as a treatment for premenstrual dysphoric disorder. J Clin Psychopharmacol. 1996;16(1):3-8.
4. Yonkers KA, Chantilis SJ. Recognition of depression in obstetric/gynecology practices. Am J Obstet Gynecol. 1995;173(2):632-638.
5. Kendell RE, Wainwright S, Hailey A, Shannon B. The influence of childbirth on psychiatric morbidity. Psychol Med. 1976;6(2):297-302.
6. Kendell RE, Rennie D, Clarke JA, Dean C. The social and obstetric correlates of psychiatric admission in the puerperium. Psychol Med. 1981;11(2):341-350.
7. O'Hara MW. Postpartum Depression. In: Alloy LB, ed. Series in Psychopathology. New York: Springer-Verlag; 1995:1-27.
8. Hamilton JA. Postpartum psychiatric syndromes. Psychiatr Clin North Am. 1989;12(1):89-103.
9. Nott PN. Psychiatric illness following childbirth in Southhampton: a case register study. Psychol Med. 1982;12:557-561.
10. Kumar R, Robson KM. A prospective study of emotional disorders in childbearing women. Br J Psychiatry. 1984;144:35-47.
11. O'Hara MW, Neunaber DJ, Zekoski EM. A prospective study of postpartum depression: prevalence, course, and predictive factors. J Abnormal Psychology. 1984;91:158-171.
12. Troutman BR, Cutrona CE. Nonpsychotic postpartum depression among adolescent mothers. J Abnorm Psychol. 1990;99(1):69-78.
13. Gregoire AJ, Kumar R, Everitt B, et al. Transdermal oestrogen for treatment of severe postnatal depression [see comments]. Lancet. 1996;347(9006):930-933.
14. Stowe ZN, Nemeroff CB. Women at risk for postpartum-onset major depression. Am J Obstet Gynecol. 1995;173(2):639-645.
15. Philipps LH, O'Hara MW. Prospective study of postpartum depression: 4 1/2-year follow-up of women and children. J Abnorm Psychol. 1991;100(2):151-155.
16. Peindl KS, Zolnik EJ, Wisner KL, Hanusa BH. Effects of postpartum psychiatric illnesses on family planning. Int J Psychiatry Med. 1995;25(3):291-300.
17. Sichel DA, Cohen LS, Robertson LM, et al. Prophylactic estrogen in recurrent postpartum affective disorder. Biol Psychiatry. 1995;38(12):814-818.
18. Research, A.f.H.C.P.a. Depression in Primary Care: Detection, Diagnosis, and Treatment. Rockville, MD: US Department of Health and Human Services; 1993.
19. Warner R, Appleby L, Whitton A, Faragher B. Demographic and obstetric risk factors for postnatal psychiatric morbidity. Br J Psychiatry. 1996;168(5):607-611.
20. Cox JL, Connor Y, Kendell RE. Prospective study of the psychiatric disorders of childbirth. Br J Psychiatry. 1982;140:111-117.
21. Harris B, Lovett L, Smith J, et al. Cardiff puerperal mood and hormone study. III. Postnatal depression at 5 to 6 weeks postpartum, and its hormonal correlates across the peripartum period. Br J Psychiatry. 1996;168(6):739-744.
22. Majewska MD, Harrison NL, Schwartz RD, et al. Steroid hormone metabolites are barbiturate-like modulators of the GABA receptor. Science. 1986;232(4753):1004-1007.
23. McEwen BS. Ovarian steroids have diverse effects on brain structure and function. In: Hammar GBaM, ed. The Modern Management of Menopause. New York: Parthenon Publishing; 1993:269-278.
24. Wisner KL, Perel JM, Findling RL. Antidepressant treatment during breast-feeding. Am J Psychiatry. 1996;153(9):1132-1137.
25. Wisner KL, Wheeler SB. Prevention of recurrent postpartum major depression. Hosp Community Psychiatry. 1994;45(12):1191-1196.
26. Dalton K. Progesterone prophylaxis used successfully in postnatal depression. The Practitioner. 1985;229:507-508.

Kathryn A. Leopold, MD, is an Assistant Professor of Obstetrics and Gynecology at Albany Medical College, New York. Lauren B. Zoschnick, MD, is a Clinical Instructor in the Department of Obstetrics and Gynecology at the University of Michigan Medical Center in Ann Arbor.