DEPRESSION AND BREASTFEEDING: WHICH POSTPARTUM PATIENTS TAKE ANTIDEPRESSANT MEDICATIONS?

Abstract

Depressed breastfeeding women may have concerns about taking antidepressant medications due to fears regarding infant exposure. We examined the clinical records of 73 breastfeeding women who sought depression treatment, to identify characteristics of those who took antidepressants. Compared to women who were not treated with pharmacotherapy, breastfeeding patients who took antidepressants had more severe symptoms, greater functional impairment, more extensive psychiatric histories, and were less likely to be involved in a committed relationship. No differences were found in age, race, or education.

INTRODUCTION

In light of the known infant health benefits of breastfeeding, both the World Health Organization and American Academy of Pediatrics endorse breast-milk as the best source of nutrients during the first 6 months of life. In addition to promoting healthy infant development, research has documented other advantages, such as maternal health benefits [Newcomb et al., 1994], opportunities for positive maternal–infant interactions [Lavelli and Poli, 1998], and economic savings [Cohen and Mrtek, 1994]. Not surprisingly, breastfeeding is increasingly viewed to be an optimal parenting choice, and the majority of pregnant women in the United States report plans to breastfeed rather than formula feed their infants following delivery.

Although breastfeeding may seem an optimal choice, the decision to breastfeed is considerably more complicated for women who suffer from major depression during the postpartum period. Currently, the US Food and Drug Administration does not approve any antidepressant medication for use during lactation. Although short-term outcome studies suggest sertraline and other selective serotonin reuptake inhibitors are likely to be safe options [Weissman et al., 2004], long-term safety data are extremely limited [Burt et al., 2001]. The availability of safe, effective, and acceptable treatments for maternal depression is critical given that the childbearing years are a time of increased risk for depression, and approximately 10–15% of women suffer a major depressive episode following delivery. A large body of literature has documented the adverse consequences associated with postpartum depression (PPD) for mothers and infants, underscoring the importance of adequate and timely treatment.

Thus, while many believe breastfeeding is a preferable choice, depressed postpartum women, and clinicians who care for them, are faced with complex decisions. It has been increasingly argued that the concept of “infant exposure” should not only include potential adverse effects from psychotropic medications transmitted via breastmilk, but also exposure to maternal depression itself, including biological, psychological, and functional changes [Newport et al., 2001]. Because the relative risk associated with each type of exposure remains unclear, guidelines have been suggested to help weigh pros and cons of pharmacotherapy during pregnancy and lactation [Altshuler et al., 2001].

Although the complexity of decisions regarding pharmacotherapy and breastfeeding is often acknowledged [e.g., Sit and Wisner, 2005], the treatment choices that postpartum patients actually make, and factors influencing those decisions, have rarely been examined. The fact that many women with PPD do not seek psychiatric care, even when faced with serious symptoms and impairment [Whitton et al., 1996], and recent findings suggesting that women may be reluctant to take antidepressants if breastfeeding [e.g., Chabrol et al., 2004; Pearlstein et al., 2006], highlight our need to better understand the acceptability of various PPD treatments and clarify how breastfeeding may influence whether and when treatment is pursued. The aim of this report is to identify some characteristics of postpartum breastfeeding patients who pursue pharmacologic treatment when they become depressed. We hypothesized that depressed women who take antidepressants, in spite of their breastfeeding status, are likely to be suffering from greater symptomatic and functional impairment compared to those who do not take antidepressants.

METHOD

Data were collected as part of a retrospective study of 500 perinatal psychiatric patients treated either at a psychiatric day hospital program or at an outpatient behavioral health clinic, both located in Providence, Rhode Island. A full description of the original study is published elsewhere [Battle et al., 2006]. Briefly, following Institutional Review Board approval, trained research staff reviewed clinical charts of pregnant and postpartum women who sought psychiatric treatment at the two facilities during a 4-year time period. Clinical data were originally recorded in the chart by mental health professionals at the facilities, including psychiatrists, psychologists, psychiatric nurses, and social workers, and this information was obtained via clinical interview and observation. The following data were abstracted from patient records: demographics, psychiatric diagnoses, Global Assessment of Functioning (GAF) ratings, and information regarding prior psychiatric history. Most records also included the Edinburgh Postnatal Depression Scale (EPDS; Cox et al. [1987]) administered at intake. The EPDS is a self-report screening measure of depressive symptoms that has been validated against clinician-administered assessments; a score above 12 signifies possible depression. Each chart was reviewed by two raters to promote accuracy of information obtained. The present analyses include only those patients who were postpartum, breastfeeding, and diagnosed with major depressive disorder (MDD) (N = 73). The majority of these women (97%) were day hospital patients. Although all patients in this sample were able to discuss pharmacotherapy options with a psychiatrist familiar with prescribing antidepressants to pregnant and lactating women, the content of those conversations and degree of encouragement to pursue medication was not known.

RESULTS

Patient age ranged from 17 to 41 years, averaging 30.2 years (SD = 5.32). The majority of women were White, non-Hispanic (71.4%), and the largest ethnic/ racial minority group was Hispanic women (20.0%). About three-fourths of the sample (73.4%) had attained some college education. Most women (71.2%) were married or in a committed relationship; however a sizable minority were divorced, separated, or otherwise not in a relationship (28.8%). At intake, women ranged from 1 to 36 weeks postpartum, with an average of 8.7 weeks. EPDS scores reflected symptoms in the clinical range (x = 19.64, SD = 4.84). Functional impairment, as measured by the clinician-rated GAF, ranged from 35 to 60, with an average of 52.3 (SD = 4.9). Forty-five women (61.6%) chose to take antidepressants during their treatment course, most (82%) taking selective serotonin reuptake inhibitors.

Table 1 compares characteristics of breastfeeding patients who took antidepressants with those who did not. Women who took antidepressants reported significantly more depressive symptoms and greater functional impairment, compared to patients who did not take medication. In addition, patients who took antidepressants were significantly more likely to have a diagnosis of severe (versus mild/moderate) depression, more likely to report a psychiatric history during their intake interview, and more likely to have had previous experience with psychotropic medications. There were also two non-significant trends: women taking antidepressants were more likely to be diagnosed with recurrent depression (χ² = 3.7, p = .06) and were slightly more weeks postpartum ((t) = −1.67, p = .09). Finally, women who took antidepressants were significantly less likely to be in a committed relationship (χ² = 4.7, p<.05); no differences were found in parity, age, race, and educational level.

TABLE 1.

Comparison of breastfeeding MDD patients who took antidepressants with those who did not take antidepressants

	Took antidepressants (n = 45)		Did not take antidepressants (n = 28)
Functioning and symptoms	Range	x (SD)	Range	x (SD)	t (df)
GAF	35–60	51.1 (5.4)	35–80	54.08 (3.7)	2.46 (61)*
EPDS	11–29	20.7 (4.9)	14–26	17.53 (3.9)	−2.42 (56)*
Depression recurrence and severity	n	%	n	%	χ2 (df)
Single episode MDD	17	37.8	17	60.7	3.7 (1)
Recurrent MDD	28	62.2	11	39.3
Current episode mild or moderate	5	11.1	10	35.7	6.4 (1)*
Current episode severe or severe with psychosis	40	88.9	18	64.3
Psychiatric history
Reports history of psychiatric disorder	33	75.0	14	51.9	4.0 (1)*
No history of psychiatric disorder	11	25.0	13	48.1
Reports history of taking psychotropic medication	27	60.0	5	19.2	11.1 (1)**
No history of taking psychotropic medication	18	40.0	21	80.8

^*p <.05.

^**p <.01.

GAF, Global Assessment of Functioning; EPDS, Edinburgh Postnatal Depression Scale.

DISCUSSION

We found that significant differences existed between depressed breastfeeding women who take antidepressants and those who do not. Women with more functional impairment and those suffering higher levels of symptoms were more likely to pursue pharmacologic treatment for their PPD. Those taking antidepressants were also more likely to have a prior psychiatric history, and were often raising their new child without the support of a partner. Not surprisingly, breastfeeding women who opted for antidepressant treatment were more likely to have had past experiences taking psychotropic medications, a likely factor in their current treatment choice.

This study’s retrospective design limits conclusions that may be drawn, particularly because women’s reasons for pursuing a given treatment were not assessed. Also, as noted earlier, the degree to which individual participants were encouraged by their clinician to pursue antidepressant treatment was unknown. Finally, the use of clinicians’ chart diagnoses and GAF ratings, rather than standardized assessments of women’s diagnosis and functional status, represents another important limitation. Nevertheless, these findings raise interesting questions that warrant further investigation. Specifically, do breastfeeding women become more open to medication treatment when symptoms are worse? To what extent does breastfeeding status, in general, play a role in patients’ treatment decisions? Although anecdotal evidence suggests that perinatal women have unique concerns that affect decisions about pursuing treatment (e.g., worries regarding separation from baby; infant exposure), it is unknown how common such concerns are, and whether they lead some women to postpone disclosure of symptoms and put off psychiatric treatment until symptoms become more serious.

Moreover, while some research has examined clinicians’ conceptualizations of PPD [Lloyd and Hawe, 2003] and their beliefs regarding the usefulness of various forms of treatment [e.g., Buist et al., 2005], little is known, more specifically, about clinicians’ beliefs regarding the appropriateness of pharmacotherapy in the context of breastfeeding, and how such beliefs influence their prescription and referral practices. Given that definitive safety information is lacking, the nature of the individualized advice from clinician to patient regarding the appropriateness of pharmacotherapy becomes more critical. It would be useful to understand how clinicians communicate with patients about these issues, as well as how patients understand advice given to them. What recommendations are made to breastfeeding patients who are reluctant to take antidepressants, particularly if access to psychotherapy is limited? Finally, how does presence of a partner alter women’s treatment decisions? The fact that single breastfeeding women were more likely to take medications could reflect patients’ or clinicians’ views that single mothers must regain adequate functioning as quickly as possible, as sole care provider. Alternatively, some partners may voice concerns about pharmacotherapy, thus steering women toward other treatment options. Further research on PDD treatment acceptability could clarify barriers to women’s treatment-seeking, and help clinicians have more productive discussions with their patients regarding these complex issues.