The Mood,Mother,and Infant Study: Associations Between Maternal Mood in Pregnancy and Breastfeeding Outcome

Abstract

Purpose:

We sought to determine the role of depression and anxiety in breastfeeding cessation.

Materials and Methods:

Participants underwent a baseline visit with a structured clinical interview in the third trimester of pregnancy. Monthly phone interviews assessed current mood symptoms and infant feeding status. We assessed the association between baseline mood and infant feeding outcomes using Cox proportional hazards regression, adjusting for infant feeding intention and sociodemographic confounders.

Results:

We enrolled 222 mother–infant dyads in late pregnancy, of whom 206 completed assessments through 12 months postpartum. We enriched our study with symptomatic women by enrolling 87 women with current depression or anxiety (Current), 64 women with a history of depression or anxiety (Past), and 71 women with no psychiatric history (Never). In multivariable-adjusted analyses, baseline diagnosis was not associated with breastfeeding outcome, but baseline symptoms of depression (Beck Depression Inventory ≥11) or anxiety (Spielberger State Anxiety ≥40) were associated with earlier introduction of formula (depression: adj hazard ratio [HR] 1.52, 95% confidence interval [CI] 1.01–2.30; anxiety: 1.70, 95% CI 1.01–2.87); and any cessation of breastfeeding (depression: adj HR 2.02, 95% CI 1.23–3.31; anxiety: 1.83, 95% CI 1.00–3.33), as were depression symptoms among women who were being treated with antidepressants, compared with untreated asymptomatic women (formula: adj HR 2.27, 95% CI 1.29–4.02; cessation: 2.32, 95% CI 1.17–4.61). History of childhood trauma (adj HR 1.34, 95% CI 1.12–1.61), disordered eating symptoms (adj HR 1.22, 95% CI 1.02–1.46), and poor sleep quality in pregnancy (adj HR 1.32, 95% CI 1.09–1.60) were independently associated with earlier introduction of formula.

Conclusions:

Baseline mood symptoms were independently associated with earlier formula introduction and cessation of breastfeeding. History of childhood trauma, disordered eating symptoms and poor sleep quality were associated with earlier formula introduction. Targeted support may enable women with these symptoms to achieve their feeding goals.

Introduction

In the weeks after childbirth, mother and infant navigate a complex neuroendocrine and behavioral transition. The mother must establish lactation and care for her neonate while coping with the loss of placental hormones, and the infant must learn to feed and elicit maternal care, relying upon the mother to regulate emotional responses to stressful stimuli.¹ Unsuccessful navigation of this transition may manifest as perinatal depression (PND), which affects more than 400,000 mothers in the United States each year.² This condition typically presents as a depressive episode, and may include anxiety symptoms during pregnancy or the first 3 months postpartum.³ Women with PND may also have suicidal ideation and/or fears of hurting the newborn.^4,5 In addition, PND is associated with reduced maternal sensitivity,⁶ which may adversely affect development of infant emotional regulation and attachment,^7–9 even after mothers have received treatment for depressed mood.¹⁰ Insecure attachment, in turn, increases risk of psychiatric disease in the child.^6,11 Thus, PND is a common, morbid condition that adversely affects the health of both mother and child.

Recent evidence implicates dysregulation of hypothalamic-pituitary-adrenal axis activity, as well as genetic and epigenetic mechanisms, in the etiology of PND.^12,13 In addition, reductions in the neuropeptide oxytocin may contribute to depression, both in the general population¹⁴ and in the perinatal period.^15,16 The central role of oxytocin in lactation and the strong association between failed lactation and PND further support this hypothesis.¹⁷

We sought to define the role of oxytocin and stress reactivity in the psychobiology of PND and impaired dyadic development, indexed by maternal sensitivity, infant emotional regulation, and insecure attachment. In this study, we describe the study design, methods, and baseline characteristics of the Mood, Mother, and Infant study cohort, and we report the association between maternal depression and anxiety symptoms and subsequent breastfeeding outcome. We hypothesized that depression and anxiety symptoms in the prenatal period would be associated with reduced duration of exclusive and any breastfeeding.

Materials and Methods

The Mood, Mother, and Infant study

We conducted a longitudinal study of mother–infant dyads from late pregnancy and through 12 months postpartum. To ensure at least 200 dyads for the 12-month visit, we recruited 222 mothers. Participants underwent a structured clinical interview at enrollment, and we enriched our study with symptomatic women by enrolling 87 women with current depression or anxiety (Current), 64 women with a history of depression or anxiety (Past), and 71 women with no psychiatric history (Never).

Recruitment of participants

We recruited women from prenatal and psychiatric clinics affiliated with the University of North Carolina Hospital, as well as through e-mail and social media. Eligible women presenting for prenatal care were approached by research assistants and invited to participate. In addition, fliers, posters, e-mails, and social media posts invited prospective participants to compete an online eligibility questionnaire or contact the study coordinator by phone. A total of 617 women completed the online eligibility questionnaire, and 362 women underwent phone screening. We enrolled about 5 women per month from May, 2013 through April 2017, for a total sample of 222 women.

Eligibility criteria

Inclusion criteria are (1) Singleton pregnancy; (2) intention to breastfeed; (3) intention to remain within 40 miles of UNC through infant's first birthday; and (4) ability to communicate in English. Exclusion criteria are (1) maternal diagnosis of Axis I disorders other than unipolar depression or anxiety disorders. We excluded women with a history of bipolar disorder, given their increased risk of postpartum psychosis. (2) Current substance use (Tobacco, alcohol, and illicits); (3) major congenital anomaly, Neonatal Intensive Care Unit admission >12 hours, and perinatal death; (4) chronic medication/medical condition contraindicated for breastfeeding; and (5) current use of tricyclic antidepressants, which alter cortisol¹⁸ and heart rate variability.^19,20

Study assessments

Demographics and medical history

Baseline demographic information was collected at the baseline interview, along with history of pregnancies and outcomes, pregnancy complications, and medications.

Maternal medications and health

Maternal medications were assessed at enrollment and monthly from 1 to 12 months postpartum. Starting at 1 month after birth, participants were asked about contraception use, dose and frequency of other medications, as well as current participation in psychotherapy, use of cigarettes or e-cigarettes, and any hospital admissions for mother or infant.

Infant feeding intention

We measured feeding intention using the Infant Feeding Intentions Scale.²¹ This instrument uses Likert-scale questions to quantify the strength of maternal intention to initiate breastfeeding and sustain exclusive breastfeeding through 6 months postpartum. In a validation study, each one point increase in infant feeding intention (IFI) score was associated with 13.7% lower hazards of formula-feeding at 30 days postpartum.

Circumstances of pregnancy

The circumstances of pregnancy scale was developed based on qualitative interviews with 47 women in the United Kingdom regarding pregnancy circumstances.²² Six domains were identified, including contraceptive use, desire for motherhood, expressed intentions, personal circumstances/timing, partner influences, and preconceptual preparations. Each domain is scored a 0, 1, or 2, with higher scores reflecting stronger pregnancy intention (total score range 0–12).

Maternal psychiatric history

At the baseline visit, subjects underwent a full psychiatric structured interview, including the Structured Clinical Interview (SCID-NP) for DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition).²³ Women were classified as having a current diagnosis of depression if they met criteria for current Major Depression Disorder, Dysthymic Disorder, Depression Disorder NOS, or Postpartum Depression. Women were classified as having a current diagnosis of anxiety if they met criteria for current Panic Disorder, Agoraphobia without history of Panic, Social Phobia, Specific Phobia, Obsessive Compulsive Disorder, Post-Traumatic Stress Disorder, or Generalized Anxiety Disorder. Women who were receiving active treatment for one of these disorders (pharmacotherapy or psychotherapy) were considered to have “Current” depression or anxiety. Women who had a history of any of these disorders but were asymptomatic and not being actively treated were classified as having a “Past” diagnosis. One woman with a history of PND but no symptoms during pregnancy had been prescribed an antidepressant before enrollment to increase her milk supply; as she was not being treated for depression or anxiety, she was classified in the “Past” group. Women with no current or past diagnoses of these disorders were classified in the “Never” group.

Edinburgh Postnatal Depression Scale

We measured maternal mood monthly using the Edinburgh Postpartum Depression Scale (EPDS), a validated assessment of perinatal mood that can be administered by phone.²⁴ A cutoff score of ≥10 was used to define mild depression, and a cutoff of ≥13 was used to define major depression.²⁵

Spielberger State/Trait Anxiety Inventory

We quantified anxiety symptoms with the Spielberger State-Trait Anxiety Inventory (Spielberger 1983). The trait inventory provides a stable measure of anxiety, whereas the state inventory captures perceived stress “right now.” We used a cutoff of >40 for state anxiety to define women as symptomatic.²⁶

Beck Depression Inventory-II

We quantified depressive symptoms with the Beck Depression Inventory-II (BDI-II).²⁷ We defined mild depression symptoms as BDI-II ≥11.²⁸

Pittsburgh Sleep Quality Index

We measured sleep quality with the Pittsburgh Sleep Quality Index,²⁹ which measures the seven domains of sleep quality, latency, duration, habitual efficiency, disturbance, use of sleeping medication, and daytime dysfunction, each on a 0–3 scale. Higher scores reflect worse sleep quality. A global score >5 provides a sensitive and specific measure of poor sleep quality.

Coping Strategies Questionnaire

We assessed maternal pain coping strategies using the Catastrophizing subscale of the Coping Strategies Questionnaire.³⁰ Higher catastrophizing scores are associated with greater affective pain scores and greater brain activity with a standardized pain stimulus.³¹

Eating Disorder Examination Questionnaire 4

The Eating Disorder Examination Questionnaire³² is a 36-item questionnaire that quantifies symptoms of disordered eating and body image in the prior 4 weeks, resulting in 4 subscales of restraint, eating concern, weight concern, and shape concern, as well as a global scale that is the mean of the four subscales.

Maternal Antenatal Attachment Scale

We quantified maternal feelings toward the fetus using the Maternal Antenatal Attachment Scale.³³ This 19-item instrument quantifies 2 factors: quality of attachment (positive/ambivalent) and intensity of preoccupation with the fetus, indexed by time spent thinking about the fetus (high/low).

Adult attachment style

At baseline, participants completed the Experience in Close Relationships Scale-Short Form,³⁴ which quantifies adult relationship style across dimensions of anxiety and avoidance.

Experience of being parented

We assessed the participant's experience of being parented before age 16 using the Parental Bonding Instrument, which quantifies dimensions of maternal and paternal care and overprotection. This instrument has shown retest coefficients >0.6 over 20-year follow-up.³⁵

Childhood Trauma Questionnaire

We used the 28-item instrument Childhood Trauma Questionnaire³⁶ to assess childhood abuse (physical, emotion, and sexual) and neglect (physical and emotional). For analyses, childhood trauma was dichotomized as “moderate/severe” or “<moderate/severe,” using previously established thresholds for abuse (physical = 10; emotional = 13; and sexual = 8) and neglect (physical = 10; emotional = 15).^36,37

Partner conflict

We quantified partner conflict with the Conflict Tactics Scale Form R subscales for verbal aggression and violence.³⁸ This instrument includes 19 statements regarding different behaviors that partners might engage in an argument, ranging from “Discussed an issue calmly” to “Used a knife or fired a gun.” Items were summed to create three scores for conflict tactics for the respondent and for the spouse or partner: Reasoning, Verbal Aggression, and Violence.

Medical Outcomes Study Social Support Survey

We quantify social support using the Medical Outcomes Study Social (MOSS) Support Survey.³⁹ The MOSS includes subscales of emotional support, tangible support, positive interaction, and affection, using a 5-point Likert scale (0 = None of the time, 4 = All of the time).

Breastfeeding duration and intensity

We assessed breastfeeding outcome with instruments adapted from the Infant Feeding Practices Survey II (IFPSII).⁴⁰ One-week recall of infant feeding, conducted via monthly interview, was used to calculate breastfeeding intensity.⁴¹ Participants reported the number of days per week that they were with their baby, and the number of days they were away from their baby.

For both “with” and “away” days, mothers were then asked to report the number of times per day or per week that the infant was fed at breast, fed expressed milk, fed donor milk, fed formula, or fed solid foods. Total feeds per week was calculated as (at breast + expressed + donor + formula + solid), and intensities were calculated using total feeds in the denominator, including breastfeeding intensity (at breast)/total, breast milk feeding (at breast + expressed + donor)/total, mother's milk feeding (at breast + expressed)/total, and pumped milk feeding (expressed)/total. Each month, women were asked if they had stopped breastfeeding or introduced formula since the last study contact and, if so, the infant's age when this occurred. These data were used to calculate time to introduction of formula and to stopping breastfeeding.

Postnatal follow-up

Beginning at 1 month after birth, we contacted participants monthly via phone interviews to maintain contact and collect information on current feeding, mood, and medications (Supplementary Table S1). At the 1 month interview, participants also reported on details of their birth and early postpartum experience. Phone interviews were scheduled at the participant's convenience and conducted by the study coordinator. Study visits at enrollment and at 2, 6, and 12 months postpartum occurred at the UNC Mother–Infant Biobehavioral Laboratory, during which maternal blood and hair samples and infant saliva samples were collected for hormonal assays (Supplementary Table S2). For details of study visit protocols, see Supplementary Data, Additional Study Instruments and Visit Procedures.

Analysis plan

To characterize our study cohort, we compared baseline measures among women with current, past, or no history of depression or anxiety using chi square tests of analysis of variance, as appropriate. To quantify the association between baseline mood and breastfeeding outcome, we used Cox proportional hazards regression to model the time to first introduction of formula and to cessation of breastfeeding. To select sociodemographic confounders for inclusion in our adjusted models, we modeled the association between infant feeding intention score and outcome, adjusting for sociodemographic confounders. Sociodemographic confounders that were not significant in models for either exclusive or any breastfeeding (type 3 p-value >0.10) were removed from the model by backward elimination. The resulting set of confounders was used for adjusted analyses of associations between baseline mood and breastfeeding outcome.

In an exploratory analysis, we further quantified the extent to which a one standard deviation change in baseline psychometric measures was associated with breastfeeding outcome in unadjusted models and in models adjusted for baseline depression symptoms (BDI <11 versus BDI ≥11), infant feeding intention and sociodemographic confounders. p-Values <0.05 were considered statistically significant. All analyses were performed using SAS 9.4 (Cary, NC).

Ethics, consent, and permissions

The Mood, Mother, and Infant study was approved by the University of North Carolina Institutional Review Board (IRB 12-2016), and each participant provided written informed consent.

Results

In Figure 1, we present a flow diagram of eligible dyads according to STROBE guidelines.

FIG. 1.

Study participant flow diagram.

Sociodemographic characteristics and baseline psychometric measures of enrolled participants are presented in Tables 1 and 2. Our study sample is predominantly white (73%), married (79.7%), and educated, with 45.5% having completed a postgraduate degree (Table 1). During recruitment, we sought to frequency match our never, past, and current depression/anxiety diagnosis populations by race-ethnicity, parity, and marital status, and as intended, these characteristics were similar among the three groups (Supplementary Table S3). We found no differences by risk level in pregnancy intendedness or breastfeeding intention. As expected, depression symptoms were associated with more adverse scores on baseline psychometric measures (Table 2).

Table 1.

Sample Demographics, n (%) or Mean (Standard Deviation), by Baseline Depression Symptoms, Indexed by Beck Depression Inventory-II ≥11

	Study sample	BDI <11	BDI ≥11	p
N	222	151	72
Age, years	31.3 (4.82)	31.7 (4.55)	30.4 (5.28)	0.07
Race/Ethnicity				0.41
White	162 (73.0)	111 (73.5)	51 (71.8)
Black or African American	24 (10.8)	14 (9.3)	10 (14.1)
Hispanic	17 (7.7)	14 (9.3)	3 (4.2)
Other/Mixed	19 (8.6)	12 (7.9)	7 (9.9)
Marital status				0.03
Married	177 (79.7)	126 (83.4)	51 (71.8)
Living with partner	22 (9.9)	15 (9.9)	7 (9.9)
Not married or living with partner	23 (10.4)	10 (6.6)	13 (18.3)
Education				0.02
<4 years college graduate	55 (24.8)	32 (21.2)	23 (32.4)
Graduated 4 years college	66 (29.7)	41 (27.2)	25 (35.2)
Postgraduate	101 (45.5)	78 (51.7)	23 (32.4)
Employed during pregnancy	179 (82.9)	133 (89.3)	46 (68.7)	0.0002
Household income				0.18
Less than $30,000	43 (19.5)	23 (15.3)	20 (28.6)
$30,000–59,999	44 (20.0)	31 (20.7)	13 (18.6)
$60,000–79,000	49 (22.3)	33 (22.0)	16 (22.9)
$80,000–99,000	31 (14.1)	24 (16.0)	7 (10.0)
$100,000 or above	53 (24.1)	39 (26.0)	14 (20.0)
Insurance				0.17
None (self-pay)	6 (2.7)	5 (3.3)	1 (1.4)
Government (Medicare/Medicaid)	55 (24.8)	31 (20.5)	24 (33.8)
TRICARE	4 (1.8)	3 (2.0)	1 (1.4)
Private	157 (70.7)	112 (74.2)	45 (63.4)
Parity				0.003
0	99 (44.6)	78 (51.7)	21 (29.6)
1	89 (40.1)	56 (37.1)	33 (46.5)
2 or more	34 (15.3)	17 (11.3)	17 (23.9)
History of early pregnancy loss	62 (27.9)	39 (25.8)	23 (32.4)	0.31
Mode of delivery				0.13
Spontaneous vaginal birth	169 (78.2)	116 (77.9)	53 (79.1)
Operative vaginal birth	8 (3.7)	8 (5.4)	0 (0.0)
Cesarean birth	39 (18.1)	25 (16.8)	14 (20.9)
Using antidepressant at baseline	40 (18.0)	20 (13.2)	20 (28.2)	0.007
High infant feeding intention	125 (56.3)	87 (57.6)	38 (53.5)	0.57

BDI-II, Beck Depression Inventory-II.

Table 2.

Baseline Psychometric Measures, Mean (Standard Deviation), by Depression Symptoms, Indexed by Beck Depression Inventory-II ≥11

Measure	Study sample	BDI <11	BDI ≥11	p
BDI-II	9.27 (7.15)	5.55 (2.73)	17.2 (7.23)	<0.0001
EPDS	5.23 (4.33)	3.32 (2.76)	9.30 (4.25)	<0.0001
STAI State Anxiety	28.0 (8.25)	25.0 (5.23)	34.3 (9.86)	<0.0001
STAI Trait Anxiety	34.3 (10.5)	29.5 (6.23)	44.5 (10.4)	<0.0001
Pregnancy intendedness score	9.96 (2.87)	10.5 (2.33)	8.82 (3.53)	<0.0001
CSQ catastrophizing score	4.04 (5.06)	2.69 (3.69)	6.90 (6.28)	<0.0001
MAAS total score	76.6 (6.27)	77.3 (5.88)	75.2 (6.86)	0.02
Childhood trauma score	35.8 (11.9)	33.7 (9.48)	40.2 (15.1)	0.0002
Parental bonding instrument
Maternal care	27.2 (8.19)	28.4 (7.44)	24.8 (9.19)	0.002
Maternal overprotection	13.3 (8.66)	11.9 (7.70)	16.1 (9.88)	0.0006
Paternal care	25.5 (8.35)	26.7 (8.02)	22.7 (8.45)	0.001
Paternal overprotection	11.3 (7.78)	10.4 (7.31)	13.2 (8.48)	0.01
Experiences in close relationships
Anxiety	18.5 (6.93)	16.9 (6.64)	21.8 (6.33)	<0.0001
Avoidance	11.0 (5.71)	9.34 (4.17)	14.4 (6.93)	<0.0001
EDE global score	0.61 (0.75)	0.43 (0.54)	0.98 (0.98)	<0.0001
PSQI total score	5.92 (3.14)	5.05 (2.76)	7.81 (3.10)	<0.0001
MOSS mean score	3.48 (0.57)	3.59 (0.49)	3.23 (0.64)	<0.0001
Conflict Tactics Scale
Partner reasoning	2.75 (1.05)	2.72 (0.98)	2.81 (1.19)	0.53
Partner verbal aggression	0.85 (0.94)	0.74 (0.91)	1.10 (0.95)	0.007
Partner violence	0.03 (0.21)	0.02 (0.20)	0.06 (0.23)	0.26
Mother reasoning	2.94 (1.03)	2.90 (1.01)	3.01 (1.06)	0.47
Mother verbal aggression	0.93 (0.84)	0.77 (0.75)	1.26 (0.93)	<0.0001
Mother violence	0.02 (0.11)	0.01 (0.04)	0.06 (0.18)	0.001

EPDS, Edinburgh Postpartum Depression Scale; MOSS, Medical Outcomes Study Social; STAI, State-Trait Anxiety Inventory.

Overall infant feeding intention in our sample was high; of the 222 women enrolled, 125 (56.3%) had an infant feeding intention score of 16, the maximum measure. We therefore dichotomized infant feeding intention as 16 or <16. Less than maximum IFI score was strongly associated with cessation of both any breastfeeding (unadj hazard ratio [HR] 2.72, 95% confidence interval [CI] 1.74–4.25, adj HR 3.53, 95% CI 1.08–5.99), and introduction of formula (unadj HR 2.23, 95% CI 1.60–3.13, adj HR 2.87, 95% CI 1.97–4.18). Sociodemographic factors that were predictors of infant feeding outcome independent of intention were marital status, race/ethnicity, education, employment in third trimester, income, parity, and insurance type. These covariates were included in multivariable adjusted models.

In proportional hazards models, baseline psychiatric diagnosis by structured clinical interview was not associated with earlier use of formula or cessation of breastfeeding (Table 3); however, both depression and anxiety symptoms were associated (BDI ≥11 adj HR formula introduction 1.52, 95% CI 1.01–2.30; adj HR cessation: 2.02, 95% CI 1.23–3.31; Spielberger State ≥40 adj HR formula introduction 1.70, 95% CI 1.01–2.87; adj HR cessation 1.83, 95% CI 1.00–3.33). When we modeled the association between BDI ≥11 and treatment with psychiatric medication, women with symptoms despite treatment had a higher risk than asymptomatic, untreated women for both formula introduction (adj HR 2.27, 95% CI 1.29–4.02), and breastfeeding cessation (adj HR 2.32, 95% CI 1.17–4.61).

Table 3.

Associations Between Infant Feeding Intention, Baseline Depression, and Anxiety Measures and Duration of Any and Exclusive Breastfeeding, Hazard Ratio, and 95% Confidence Interval

	Introduction of formula		Cessation of breastfeeding
	Unadj, HR (95% CI)	Adj ^a HR (95% CI)	Unadj, HR (95% CI)	Adj ^a HR (95% CI)
Infant feeding intention score
16	1.0 (ref)	1.0 (ref)	1.0 (ref)	1.0 (ref)
<16	2.72 (1.74–4.25)	3.53 (2.08–5.99)	2.23 (1.60–3.13)	2.87 (1.97–4.18)
Risk level assessed by structured clinical interview
No depression or anxiety	1.0 (ref)	1.0 (ref)	1.0 (ref)	1.0 (ref)
Past depression or anxiety	1.26 (0.82–1.95)	1.53 (0.95–2.48)	1.33 (0.77–2.31)	1.41 (0.76–2.61)
Current depression or anxiety	1.32 (0.88–1.98)	1.34 (0.85–2.10)	1.19 (0.70–2.01)	1.21 (0.67–2.21)
Depression symptoms
BDI <11	1.0 (ref)	1.0 (ref)	1.0 (ref)	1.0 (ref)
BDI ≥11	1.57 (1.11–2.22)	1.52 (1.01–2.30)	1.75 (1.13–2.71)	2.02 (1.23–3.31)
Anxiety symptoms
STAI State <40	1.0 (ref)	1.0 (ref)	1.0 (ref)	1.0 (ref)
STAI State ≥40	1.84 (1.13–2.98)	1.70 (1.01–2.87)	2.06 (1.18–3.61)	1.83 (1.00–3.33)
Symptoms and medication
BDI <11, no medication	1.0 (ref)	1.0 (ref)	1.0 (ref)	1.0 (ref)
BDI <11, on medication	1.25 (0.76–2.04)	1.38 (0.82–2.34)	0.94 (0.47–1.88)	0.98 (0.48–2.03)
BDI ≥11, no medication	1.56 (1.01–2.42)	1.28 (0.78–2.11)	1.80 (1.05–3.07)	1.82 (0.99–3.37)
BDI ≥11, on medication	1.78 (1.10–2.89)	2.27 (1.29–4.02)	1.65 (0.90–3.04)	2.32 (1.17–4.61)

Bold indicates a statistically significant association.

Adjusted or marital status, race/ethnicity, education, employment in third trimester, income, parity, insurance, and infant feeding intention.

CI, confidence interval; HR, hazard ratio.

In an exploratory analysis, we evaluated the association between other psychometric measures and breastfeeding outcomes (Table 4). We found that childhood trauma, eating disorder symptoms, and poor sleep quality were associated with earlier introduction of formula in models adjusting for depression symptoms, feeding intention, and sociodemographic covariates. In adjusted models, none of the psychometric measures evaluated was independently associated with breastfeeding cessation.

Table 4.

Association of Psychometric Measures with Breastfeeding Outcome, Hazard Ratio, and 95% Confidence Interval per Standard Deviation Change

	Introduction of formula		Cessation of breastfeeding
	Unadj, HR (95% CI)	Adj ^a HR (95% CI)	Unadj, HR (95% CI)	Adj ^a HR (95% CI)
STAI Trait	1.12 (0.94–1.32)	1.02 (0.80–1.30)	1.16 (0.95–1.42)	0.98 (0.72–1.33)
Pregnancy intendedness score	0.81 (0.69–0.95)	1.11 (0.88–1.40)	0.78 (0.64–0.95)	0.96 (0.74–1.24)
CSQ catastrophizing score	1.17 (0.99–1.38)	1.01 (0.82–1.24)	1.20 (0.99–1.46)	0.96 (0.75–1.25)
Maternal Antenatal Attachment Scale	0.96 (0.81–1.15)	0.98 (0.80–1.20)	1.08 (0.86–1.36)	1.19 (0.91–1.56)
Childhood trauma score	1.42 (1.21–1.66)	1.34 (1.12–1.61)	1.24 (1.02–1.51)	1.14 (0.89–1.46)
Parental bonding instrument
Maternal care	0.80 (0.68–0.93)	0.86 (0.73–1.02)	0.85 (0.70–1.04)	0.99 (0.79–1.24)
Paternal care	0.87 (0.73–1.04)	0.88 (0.73–1.07)	0.86 (0.68–1.08)	0.91 (0.70–1.18)
Maternal overprotection	1.28 (1.10–1.50)	1.14 (0.96–1.36)	1.30 (1.07–1.58)	1.15 (0.91–1.44)
Paternal overprotection	1.16 (0.98–1.37)	1.11 (0.92–1.35)	1.17 (0.94–1.46)	1.16 (0.89–1.50)
Experiences in close relationships
Anxiety	1.18 (1.00–1.39)	1.06 (0.87–1.30)	1.10 (0.89–1.35)	0.94 (0.73–1.22)
Avoidance	1.24 (1.07–1.45)	1.07 (0.88–1.31)	1.03 (0.84–1.26)	0.87 (0.65–1.16)
Eating disorder exam global score	1.33 (1.14–1.55)	1.22 (1.02–1.46)	1.30 (1.08–1.57)	1.13 (0.90–1.41)
Pittsburgh sleep quality score	1.36 (1.15–1.62)	1.32 (1.09–1.60)	1.34 (1.09–1.66)	1.22 (0.94–1.59)
Medical Outcomes Support Scale	0.96 (0.83–1.13)	1.13 (0.93–1.36)	1.08 (0.87–1.33)	1.25 (0.97–1.61)
Conflict Tactics Scale
Partner reasoning	0.89 (0.74–1.08)	1.01 (0.83–1.23)	0.81 (0.65–1.02)	0.85 (0.67–1.09)
Partner violence	0.97 (0.83–1.14)	0.98 (0.80–1.19)	0.79 (0.48–1.30)	0.73 (0.36–1.50)
Partner verbal aggression	0.93 (0.78–1.11)	0.83 (0.68–1.03)	0.96 (0.77–1.20)	0.83 (0.63–1.08)
Mother reasoning	0.86 (0.72–1.03)	1.00 (0.83–1.21)	0.76 (0.61–0.95)	0.82 (0.65–1.04)
Mother verbal aggression	1.05 (0.89–1.23)	0.93 (0.77–1.12)	1.02 (0.82–1.26)	0.89 (0.70–1.13)
Mother violence	1.03 (0.88–1.21)	0.86 (0.72–1.04)	0.91 (0.70–1.19)	0.79 (0.59–1.05)

Bold indicates p < 0.05.

Adjusted for depression symptoms at baseline, indexed by BDI ≥11, marital status, race/ethnicity, education, employment in third trimester, income, parity, insurance, and infant feeding intention.

Discussion

Consistent with our hypothesis, we found that depression and anxiety symptoms were associated with earlier introduction of formula and earlier cessation of breastfeeding. Women with persistent depression symptoms while being treated with antidepressants were at the highest risk. In addition, childhood trauma history, eating disorder symptoms, and poor sleep quality were associated with earlier introduction of formula.

Our findings confirm and extend earlier work on the association between depression symptoms during pregnancy and curtailed breastfeeding. While some authors have found that depression symptoms during pregnancy are associated with reduced breastfeeding intention,^42,43 others have not.⁴⁴ The preponderance of studies has found that women with antenatal depression symptoms have shorter durations of any^45–48 and exclusive^49,50 breastfeeding, although others have not found an association with continuation through 1 month⁵¹ or 3 months⁴⁴ postpartum. Antenatal anxiety has also been linked with reducing breastfeeding intensity and duration.⁵² Our findings extend this work by quantifying associations with mood disorder diagnosis, symptoms and treatment, and adjusting for antenatal breastfeeding intention using a validated instrument.²¹ We were further able to measure associations of childhood trauma, disordered eating and sleep quality with postnatal feeding outcomes. Although Prentice et al.⁵³ found that women with a history of sexual abuse were more likely to initiate breastfeeding, most studies have found that women who were maltreated in childhood are less likely to sustain breastfeeding.^54–56 Body image dissatisfaction has similarly been associated with reduced breastfeeding intention and duration,^57,58 especially among women who are overweight.⁵⁹ No prior studies to our knowledge have quantified associations between poor sleep quality during pregnancy and subsequent exclusive breastfeeding; however, in cross-sectional studies, mixed feeding has been associated with worse sleep quality than exclusive breastfeeding.^60,61 Our findings suggest that preexisting maternal sleep problems may be a risk factor for mixed feeding, rather than mixed feeding causing sleep disruption.

Strengths of our study include assessment of mood history with a structured clinical interview, comprehensive psychometric measures, and longitudinal follow-up of feeding outcomes through 12 months. However, our findings must be interpreted in the context of the study design. Because assessment of oxytocin during breastfeeding is a central component of the parent study, we enrolled women who intended to breastfeed for at least 2 months. These inclusion criteria may affect generalizability, in that associations between mood and feeding patterns may differ in women without a strong prenatal intention to breastfeed. In addition, our sample is predominantly white, highly educated, and middle to upper-middle class, and thus not subject to structural determinants of health that may impact other populations. Finally, although we enrolled 88 women with a current diagnosis of depression or anxiety, mean symptom scores on the Edinburgh Postnatal Depression Scale, BDI-II, and Spielberger State Anxiety Inventory were below clinical thresholds. We speculate that women with more severe symptoms may be reluctant to engage in an intensive longitudinal study, resulting in a cohort of women with relatively mild symptoms. Associations between well-controlled depression or anxiety and outcomes of interest may differ from those among women with more severe symptoms.

Conclusions

We have successfully recruited and retained a longitudinal cohort of mothers and infants recruited during the third trimester of pregnancy with detailed phenotyping data on mood, relationships, infant feeding, and social supports. We found that maternal symptoms of depression and anxiety during pregnancy were associated with earlier cessation of exclusive and any breastfeeding. We further found that women with a history of childhood trauma, disordered eating symptoms, and poor prenatal sleep quality were at increased risk. Targeted support in these domains may enable more women to meet their breastfeeding goals.

Footnotes

Acknowledgment

The authors are grateful to the families who participated in the Mood, Mother, and Infant study.

Disclosure Statement

A.M.S. and S.M.B. receive grant support from Janssen Research and Development, and S.M.B., receives grant support from Sage Therapeutics, Inc., awarded to the University of North Carolina (Chapel Hill, NC). These grants are outside the submitted work. The other authors have no competing interests.

Funding Information

This research was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development (R01 HD073220-01). The National Institutes of Health had no role in the design of the study and collection, analysis, and interpretation of data and in writing the article.

Supplementary Material

References

Cox

, Mills-Koonce

, Propper

, et al. Systems theory and cascades in developmental psychopathology. Dev Psychopathol, 2010; 22:497–506.

Gaynes

, Gavin

, Meltzer-Brody

, et al. Perinatal depression: Prevalence, screening accuracy, and screening outcomes. Evid Rep Technol Assess (Summ), 2005:1–8.

APA. Diagnostic and Statistical Manual of Mental Disorders, 4th ed. Washington, DC: American Psychiatric Association, 1994.

Bernstein

, Rush

, Yonkers

, et al. Symptom features of postpartum depression: Are they distinct?. Depress Anxiety, 2008; 25:20–26.

Abramowitz

, Meltzer-Brody

, Leserman

, et al. Obsessional thoughts and compulsive behaviors in a sample of women with postpartum mood symptoms. Arch Womens Ment Health, 2010; 13:523–530.

Burke

The impact of maternal depression on familial relationships. Int Rev Psychiatry, 2003; 15:243–255.

Ashman

, Dawson

Maternal depression, infant psychobiological development, and risk for depression. In: Children of Depressed Parents, Goodman

, Gotlib

, eds. Washington, DC: American Psychological Association, 2002.

Field

, Sandberg

, Garcia

, et al. Pregnancy problems, postpartum depression, and early mother and infant interactions. Dev Psychol, 1985; 21:1152–1156.

Netsi

, Evans

, Wulff

, et al. Infant outcomes following treatment of antenatal depression: Findings from a pilot randomized controlled trial. J Affect Disord, 2015; 188:252–256.

10.

Forman

, O'Hara

, Stuart

, et al. Effective treatment for postpartum depression is not sufficient to improve the developing mother-child relationship. Dev Psychopathol, 2007; 19:585–602.

11.

Beck

CT.

The effects of postpartum depression on maternal-infant interaction: A meta-analysis. Nurs Res, 1995; 44:298–304.

12.

Meltzer-Brody

New insights into perinatal depression: Pathogenesis and treatment during pregnancy and postpartum. Dialogues Clin Neurosci, 2011; 13:89–100.

13.

Cox

, Stuebe

, Pearson

, et al. Oxytocin and HPA stress axis reactivity in postpartum women. Psychoneuroendocrinology, 2015; 55:164–172.

14.

Gordon

, Zagoory-Sharon

, Schneiderman

, et al. Oxytocin and cortisol in romantically unattached young adults: Associations with bonding and psychological distress. Psychophysiology, 2008; 45:349–352.

15.

Skrundz

, Bolten

, Nast

, et al. Plasma oxytocin concentration during pregnancy is associated with development of postpartum depression. Neuropsychopharmacology, 2011; 36:1886–1893.

16.

Stuebe

, Grewen

, Meltzer-Brody

. Association between maternal mood and oxytocin response to breastfeeding. J Women's Health (Larchmt), 2013; 22:352–361.

17.

Stuebe

, Horton

, Chetwynd

, et al. Prevalence and risk factors for early, undesired weaning attributed to lactation dysfunction. J Womens Health (Larchmt), 2014; 23:404–412.

18.

Deuschle

, Hamann

, Meichel

, et al. Antidepressive treatment with amitriptyline and paroxetine: Effects on saliva cortisol concentrations. J Clin Psychopharmacol, 2003; 23:201–205.

19.

Rechlin

Decreased parameters of heart rate variation in amitriptyline treated patients: Lower parameters in melancholic depression than in neurotic depression—A biological marker?. Biol Psychiatry, 1994; 36:705–707.

20.

Roose

, Laghrissi-Thode

, Kennedy

, et al. Comparison of paroxetine and nortriptyline in depressed patients with ischemic heart disease. JAMA, 1998; 279:287–291.

21.

Nommsen-Rivers

, Dewey

. Development and validation of the infant feeding intentions scale. Matern Child Health J, 2009; 13:334–342.

22.

Barrett

, Smith

, Wellings

. Conceptualisation, development, and evaluation of a measure of unplanned pregnancy. J Epidemiol Community Health, 2004; 58:426–433.

23.

First

, Spitzer

, Gibbon

, et al. Structured Clinical Interview for DSM-IV-TR Axis I Disorders, Research Version, Non-Patient Edition (SCID-I/NP). New York: Biometrics Research, New York State Psychiatric Institute, 2002.

24.

Horowitz

, Murphy

, Gregory

, et al. A community-based screening initiative to identify mothers at risk for postpartum depression. J Obstet Gynecol Neonatal Nurs, 2011; 40:52–61.

25.

Murray

, Carothers

. The validation of the Edinburgh Post-natal Depression Scale on a community sample. Br J Psychiatry, 1990; 157:288–290.

26.

Dennis

, Coghlan

, Vigod

. Can we identify mothers at-risk for postpartum anxiety in the immediate postpartum period using the State-Trait Anxiety Inventory?. J Affect Disord, 2013; 150:1217–1220.

27.

Beck

, Steer

, Ball

, et al. Comparison of Beck Depression Inventories-IA and -II in psychiatric outpatients. J Pers Assess, 1996; 67:588–597.

28.

Wang

Y-P

, Gorenstein

. Assessment of depression in medical patients: A systematic review of the utility of the Beck Depression Inventory-II. Clinics, 2013; 68:1274–1287.

29.

Buysse

, Reynolds

3rd , Monk

, et al. The Pittsburgh Sleep Quality Index: A new instrument for psychiatric practice and research. Psychiatry Res, 1989; 28:193–213.

30.

Rosenstiel

, Keefe

. The use of coping strategies in chronic low back pain patients: Relationship to patient characteristics and current adjustment. Pain, 1983; 17:33–44.

31.

Gracely

, Geisser

, Giesecke

, et al. Pain catastrophizing and neural responses to pain among persons with fibromyalgia. Brain, 2004; 127(Pt 4):835–843.

32.

Fairburn

, Beglin

. Assessment of eating disorders: Interview or self-report questionnaire?. Int J Eat Disord, 1994; 16:363–370.

33.

Condon

JT.

The assessment of antenatal emotional attachment: Development of a questionnaire instrument. Br J Med Psychol, 1993; 66(Pt 2):167–183.

34.

Wei

, Russell

, Mallinckrodt

, et al. The Experiences in Close Relationship Scale (ECR)-Short Form: Reliability, validity, and factor structure. J Pers Assess, 2007; 88:187–204.

35.

Wilhelm

, Niven

, Parker

, et al. The stability of the Parental Bonding Instrument over a 20-year period. Psychol Med, 2005; 35:387–393.

36.

Bernstein

, Fink

Childhood Trauma Questionnaire: A Retrospective Self-Report Manual. San Antonio, TX: The Psychological Corporation, 1998.

37.

Diesel

, Bodnar

, Day

, et al. Childhood maltreatment and the risk of pre-pregnancy obesity and excessive gestational weight gain. Matern Child Nutr, 2016; 12:558–568.

38.

Straus

, Gelles

. Physical violence in American Families: Risk Factors and Adaptations to Violence in 8,145 Families. New Bruswick, NJ: Transaction, 1990.

39.

Sherbourne

, Stewart

. The MOS social support survey. Soc Sci Med, 1991; 32:705–714.

40.

Fein

, Labiner-Wolfe

, Shealy

, et al. Infant Feeding Practices Study II: Study methods. Pediatrics, 2008; 122(Suppl 2):S28–S35.

41.

, Fein

, Grummer-Strawn

. Association of breastfeeding intensity and bottle-emptying behaviors at early infancy with infants' risk for excess weight at late infancy. Pediatrics, 2008; 122(Suppl 2):S77–S84.

42.

Fairlie

, Gillman

, Rich-Edwards

High pregnancy-related anxiety and prenatal depressive symptoms as predictors of intention to breastfeed and breastfeeding initiation. J Womens Health (Larchmt), 2009.

43.

Insaf

, Fortner

, Pekow

, et al. Prenatal stress, anxiety, and depressive symptoms as predictors of intention to breastfeed among Hispanic women. J Womens Health (Larchmt), 2011; 20:1183–1192.

44.

Bogen

, Hanusa

, Moses-Kolko

, et al. Are maternal depression or symptom severity associated with breastfeeding intention or outcomes?. J Clin Psychiatry, 2010; 71:1069–1078.

45.

Pippins

, Brawarsky

, Jackson

, et al. Association of breastfeeding with maternal depressive symptoms. J Womens Health (Larchmt), 2006; 15:754–762.

46.

Hahn-Holbrook

, Haselton

, Dunkel Schetter

, et al. Does breastfeeding offer protection against maternal depressive symptomatology?: A prospective study from pregnancy to 2 years after birth. Arch Womens Ment Health, 2013; 16:411–422.

47.

Kehler

HLM

, Chaput

KHM

, Tough

SCP

. Risk factors for cessation of breastfeeding prior to six months postpartum among a community sample of women in Calgary, Alberta. Can J Public Health, 2009; 100:376–380.

48.

Yusuff

, Tang

, Binns

, et al. Antenatal Depressive Symptoms and Breastfeeding: A prospective cohort study. Breastfeed Med, 2015; 10:300–304.

49.

Cato

, Sylven

, Georgakis

, et al. Antenatal depressive symptoms and early initiation of breastfeeding in association with exclusive breastfeeding six weeks postpartum: A longitudinal population-based study. BMC Pregnancy Childbirth, 2019; 19:49.

50.

Eastwood

, Ogbo

, Hendry

, et al. The impact of antenatal depression on perinatal outcomes in Australian women. PLoS One, 2017; 12:e0169907.

51.

Chung

, McCollum

, Elo

, et al. Maternal depressive symptoms and infant health practices among low-income women. Pediatrics, 2004; 113:e523.

52.

Adedinsewo

, Fleming

, Steiner

, et al. Maternal anxiety and breastfeeding: Findings from the MAVAN (Maternal Adversity, Vulnerability and Neurodevelopment) Study. J Hum Lact, 2014; 30:102–109.

53.

Prentice

, Lu

, Lange

, et al. The association between reported childhood sexual abuse and breastfeeding initiation. J Hum Lact, 2002; 18:219–226.

54.

Coles

, Anderson

, Loxton

. Breastfeeding duration after childhood sexual abuse: An Australian cohort study. J Hum Lact, 2016; 32:Np28–Np35.

55.

Sorbo

, Lukasse

, Brantsaeter

, et al. Past and recent abuse is associated with early cessation of breast feeding: Results from a large prospective cohort in Norway. BMJ Open, 2015; 5:e009240.

56.

Ukah

, Adu

, De Silva

, et al. The impact of a history of adverse childhood experiences on breastfeeding initiation and exclusivity: Findings from a National Population Health Survey. Breastfeed Med, 2016; 11:544–550.

57.

Barnes

, Stein

, Smith

, et al. Extreme attitudes to body shape, social and psychological factors and a reluctance to breast feed. ALSPAC Study Team. Avon Longitudinal Study of Pregnancy and Childhood. J R Soc Med, 1997; 90:551–559.

58.

Brown

, Rance

, Warren

. Body image concerns during pregnancy are associated with a shorter breast feeding duration. Midwifery, 2015; 31:80–89.

59.

Bigman

, Wilkinson

, Homedes

, et al. Body image dissatisfaction, obesity and their associations with breastfeeding in Mexican women, a cross-sectional study. Matern Child Health J, 2018; 22:1815–1825.

60.

Kendall-Tackett

, Cong

, Hale

. Depression, sleep quality, and maternal well-being in postpartum women with a history of sexual assault: A comparison of breastfeeding, mixed-feeding, and formula-feeding mothers. Breastfeed Med, 2013; 8:16–22.

61.

Dorheim

, Bondevik

, Eberhard-Gran

, et al. Sleep and depression in postpartum women: A population-based study. Sleep, 2009; 32:847–855.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.11 MB