Intimate Partner Violence and Breastfeeding Behaviors: Does the Source of Breastfeeding Information Affect the Associations Between Prepregnancy or Prenatal IPV and Breastfeeding Behaviors of Women in the United States? A PRAMS 2018 Study

Abstract

Background:

Most countries do not meet World Health Organization's breastfeeding recommendations, and exposure to intimate partner violence (IPV) hinders positive breastfeeding behaviors. One in four U.S. women (43.6 million women) experiences IPV. This study aims to assess relationships between IPV, and breastfeeding initiation, duration, and early cessation among women in 42 U.S. states; and to evaluate possible modifying effect(s) of different breastfeeding information sources.

Methods:

Centers for Disease Control and Prevention's 2016–2018 Pregnancy Risk Assessment Monitoring System data (n = 105,230) were used to assess relationships between prepregnancy/prenatal IPV and breastfeeding initiation, duration, and early cessation; and modify effects of various breastfeeding information sources on study associations using multilogistic regression models.

Results:

About 1.4% of women experienced prenatal IPV with reduced odds of breastfeeding for 6 months or more (odds ratio [OR] = 0.74; 95% confidence interval = 0.58–0.94). Receiving breastfeeding information from baby's doctor modified early cessation (0.37 [0.18–0.78]) (p for interaction = 0.009) with prenatal IPV exposure. Among women exposed to prenatal IPV, breastfeeding initiation was stronger in women who received breastfeeding information from family/friends (2.46, [1.24–4.88]) (p for interaction = 0.010) or from breastfeeding support groups (3.03 [1.17–7.88]) (p for interaction = 0.023) compared to those who did not. Breastfeeding information from family/friends modified association between prepregnancy IPV and breastfeeding duration (0.67 [0.45–0.99]) (p for interaction = 0.042).

Conclusions:

Prenatal IPV is a risk factor for short-duration breastfeeding. Receiving information from doctors, nurses, support groups, and family/friends may improve breastfeeding behavior among IPV-exposed women. Interventions promoting breastfeeding information dissemination by family/friends, support groups, and doctors/nurses during hospital visits are encouraged.

Introduction

According to a 2015 survey, one in every four women (43.6 million women) in the United States experienced intimate partner violence (IPV) in their lifetime.¹ The prevalence of prenatal IPV in United States is 3% to 9%.² Prenatal IPV can affect abused women's emotional state, undermine their confidence and self-esteem,³ and foster negative breastfeeding behaviors,^4,5 reducing infant bonding⁶ and increasing breastfeeding avoidance (refusal to breastfeed or pump milk).⁷

The World Health Organization (WHO) recommends initiating breastfeeding within 1 hour of birth, and exclusive breastfeeding—on demand—as often as the child needs.^8,9 Unfortunately, most countries do not meet these breastfeeding recommendations.⁸ In high-income countries like the United States, short breastfeeding durations are common,¹⁰ and fewer than one in five babies are breastfed throughout their first year. In 2017, 84.1% of infants started breastfeeding, but only 58.3% continued after 6 months as recommended by WHO.¹¹

Despite the protective characteristics of breastfeeding, U.S. women may not fully understand its benefits.¹² Complications of juggling work/life, while nursing a baby can deter a new mother from breastfeeding—even without the added emotional and physical trauma from prepregnancy/prenatal IPV exposure.⁹ Lack of knowledge, poor social/family support, and feelings of embarrassment—which are barriers to breastfeeding in the United States¹³—are exacerbated by exposure to IPV. Breastfeeding decisions may therefore depend on personal and/or professional support, and maternal knowledge of breastfeeding benefits.¹⁴ The importance of breastfeeding information and its source (family/friends, health providers, support groups, hotlines, etc.) can never be overstated—especially as information is not always readily available/easy to understand.¹⁵

This builds on the new mother's concerns for and their desire to protect their newborn despite their exposure to IPV.^16,17 Health care providers should be a good source of breastfeeding information during pregnancy.¹⁴ However, a gap between actual clinical practice and breastfeeding benefits, and inconsistent breastfeeding assessment skills reported by health care providers in recent studies¹⁴ means that nurses and doctors may feel underutilized/ill-equipped to provide accurately adequate breastfeeding information to their patients^14,18,19 and may make referrals to lactation consultants when faced with complications.¹⁴ This means that new mothers may receive conflicting and at times, inappropriate breastfeeding information from their doctors.

Although IPV and some breastfeeding behaviors have previously been assessed, no study to our knowledge has tried to assess modifying effects of different sources of breastfeeding information (mother's doctor, pediatrician, breastfeeding hotline, family/friends, nurse/doula/midwife, or support groups) on relationships between IPV and breastfeeding behaviors.

Materials and Methods

Aim of study

This study has two goals. The first is to understand if domestic violence affects breastfeeding behaviors of women in the United States and the second is to understand how the source of breastfeeding information affects breastfeeding behaviors of women in the United States, who have been exposed to domestic abuse. We hypothesize that there is a relationship between exposure to IPV (in the 12 months before pregnancy or while pregnant) and the breastfeeding behaviors of breastfeeding initiation, breastfeeding duration, and early breastfeeding cessation in women living in 42 U.S. states.

We also hypothesize that receiving breastfeeding information from doctors, nurses, family/friends, breastfeeding support groups, and hotlines will modify the association between both types of IPV exposure and the three breastfeeding behaviors of interest. Our findings should highlight most efficient areas for health education interventions, which will improve maternity care practices provided to new mothers. Hospital practices based on evidence, and individualized support, are necessary for creating and sustaining positive breastfeeding behaviors.¹⁵ Understanding the most effective ways to provide breastfeeding information to pregnant women will help create targeted and more effective interventions to promote positive breastfeeding behavior among women in the United States.

Study design

The 2016–2018 data were obtained from the national Pregnancy Risk Assessment Monitoring System (PRAMS), an ongoing self-reported population-based surveillance system that collects state-specific data on maternal experiences before and through pregnancy using self-administered questionnaires mailed to respondents in 42 states and covers about 83% of all U.S. births. State birth certificates provided the sampling frame for sample selection and weighting, and nonresponse/noncoverage adjustments were applied to account for nonresponse bias and oversampling/undersampling characteristics of different demographics across the nation. Stratification plans were set up by each state as their population and data needs required. More information about sample design and methodology can be found elsewhere (https://www.cdc.gov/prams/methodology.htm).

Measures

The sample selection criteria were restricted to women whose baby/babies were alive at time of survey and had responded to the IPV and at least one of the breastfeeding questions of interest. About 2,880 (2.66%) missing observations for the prepregnancy/prenatal IPV questions were eliminated from the sample population of 108,110 women; thus, the final sample size for our study was 105,230 women. Breastfeeding behaviors (breastfeeding initiation, early cessation, and breastfeeding duration) and sources of breastfeeding information (from woman's doctor, baby's doctor, nurse/doula/midwife, support group, family/friends, and breastfeeding hotline) were measured using the PRAMS core questionnaire.

Prepregnancy IPV was defined as a positive response to the question: “In the 12 months before you got pregnant with your new baby, did any of the following people push, hit, slap, kick, choke, or physically hurt you in any other way?” and Prenatal IPV was defined as a positive response to the question: “During your most recent pregnancy, did any of the following people push, hit, slap, kick, choke, or physically hurt you in any other way?” Breastfeeding initiation was defined as a positive response to the question: “Did you ever breastfeed or pump breast milk to feed your new baby, even for a short period of time?”

The numerical responses to the question “How many weeks or months did you breastfeed or feed pumped milk to your baby?” allowed us to classify breastfeeding cessation into the following: “Early cessation (≤1 week)” and “Continued breastfeeding (>1 week).” Moreover, this response also allowed us to classify breastfeeding duration as follows: “Short duration (<6 months)” and “Long duration (≥6 months).”

Based on previous studies,^10,20 maternal age (“17–19 years,” “20–34 years,” and “≥35 years”), marital status (“Married” and “Other”), annual household income (“$0–$28,000,” “$28,001–$40,000,” “$40,001–$100,000,” and “$100,001+”), education level (“High school or less [<13 years]” or “More than high school [≥13 years]”), residence (Urban/Rural), race (Black, Asian, White, or Other), and ethnicity (“Hispanic” or “non-Hispanic”) were considered confounders in our study.

Breastfeeding information from the woman's doctor, family/friends, baby's doctor, breastfeeding support groups, hotlines, or nurse/doula/midwife were defined as dichotomous (Yes/No) responses to the following question: “Before or after your new baby was born, did you receive information about breastfeeding from any of the following sources?” for each source.

Statistical analysis

PRAMS 2018 dataset was accessed with Centers for Disease Control and Prevention's permission. Multivariate logistic regressions were used to estimate odds ratios (ORs)—at 95% confidence intervals (CIs)—for the association between the independent variables (prepregnancy and prenatal IPV) and dependent variables (breastfeeding initiation, early cessation, and breastfeeding duration), after adjusting for main confounders.

The interaction effects of different sources of breastfeeding information on the association between each IPV type and the different breastfeeding behaviors of interest assessed in this study were analyzed using likelihood ratio tests that compared multivariate logistic regression models with and without interaction terms. Complex Samples Modules were used for data analyses to increase generalizability. Statistical significance of two-tailed p ≤ 0.05 for all analyses was applied in this study. Adelphi university's Institutional Review Board provided IRB exemption since the cross-sectional secondary data were de-identified at source. The analyses were preformed with the program IBM SPSS Statistics for Windows, Version 26.0. Armonk, NY: IBM Corp., released in 2019.

Results

Of 105,230 women with live babies who answered both IPV questions, 1,978 (1.6%) experienced prepregnancy IPV and 1,656 (1.4%) experienced prenatal IPV. About 58,189 (82.6%) of them initiated breastfeeding, 9,881 (24.3%) experienced early cessation, and 71,023 (69.5%) breastfed for at least 6 months.

Sociodemographic characteristics of study participants

The sample consisted of about 4,816 (4.1%) teenagers between 17 and 19 years of age, 81,267 (77.6%) young adults between 20 and 34 years of age, and 19,146 (18.3%) mature adults who were 35 years of age or older. Approximately 63,009 (62.7%) were married women, 96,700 (37.9%) were lower class, 10,798 (10.7%) were working class, 23,123 (24%) were middle class, and 22,951 (27.4%) were upper class. Around 65,684 (63.7%) women had more than a high school education and 52,742 (59.6%) had partners/husbands with more than a high school education. Approximately 41,103 women (77.2%) were urban dwellers, 19,296 (14.9%) were Black, 7,061 (6%) were Asian, 60,318 (69.9%) were White, 14,924 (9.2%) were “other race,” and 10,863 (18.9%) were Hispanic.

Approximately 14,079 (13.9%) of the women had partners who were Black, 5,973 (6%) had partners who were Asian, 57,158 (71.1%) had partners who were White, and 10,534 (9%) had partners who identified as “other race.” Approximately 80,224 (77.7%) women received breastfeeding information from their doctor, 70,297 (68.9%) from their baby's doctor, 76,333 (74.7%) from their nurse/doula/midwife, 24,612 (24.1%) from breastfeeding support groups, and 11,366 (10.8%) from breastfeeding hotlines (Table 1).

Table 1.

Univariate Analysis of Women in the Study

	Total, n (%)^a
Population count (unweighted) n	105,230
IPV
Prepregnancy IPV
No	103,252 (98.4)
Yes	1,978 (1.6)
Prenatal IPV
No	103,574 (98.6)
Yes	1,656 (1.4)
Breastfeeding behaviors
Breastfeeding initiation
No	12,500 (17.4)
Yes	58,189 (82.6)
Early breastfeeding cessation
Yes (≤1 week)	9,881 (24.3)
No (>1 week)	31,557 (75.7)
Breastfeeding duration
Short duration (<6 months)	31,235 (30.5)
Long duration (≥6 months)	71,023 (69.5)
Characteristics of participants
Woman's age
Teen mom (17–19 years)	4,815 (4.1)
Young adult mom (20–34 years)	81,267 (77.6)
Matured mom (≥35 years)	19,146 (18.3)
Woman's marital status
Married	63,009 (62.7)
Other	42,119 (37.3)
Woman's annual household income
Lower class (≤28,000)	96,799 (37.9)
Working class ($28,001–$40,000)	10,798 (10.7)
Middle class ($40,001–$100,000)	23,123 (24.0)
Upper class (≤$100,001)	22,951 (27.4)
Woman's highest education level
≤High school (<13 years)	38,492 (36.3)
>High school (13 years+)	65,684 (63.7)
Woman's residence
Urban	41,103 (77.2)
Rural	17,777 (22.8)
Woman's race
Black	19,296 (14.9)
Asian	7,061 (6.0)
White	60,318 (69.9)
Other	14,924 (9.2)
Woman's ethnicity
Hispanic	19,863 (18.9)
Non-Hispanic	82,209 (81.1)
Partner's characteristics
Partner's highest education level
≤High school (<13 years)	36,951 (40.4)
>High school (13 years+)	52,742 (59.6)
Partner's race
Black	14,079 (13.9)
Asian	5,973 (6.0)
White	57,158 (71.1)
Other	10,534 (9.0)
Sources of breastfeeding information
Breastfeeding information from woman's doctor
No	21,614 (22.3)
Yes	80,224 (77.7)
Breastfeeding information from baby's doctor
No	29,743 (31.1)
Yes	70,297 (68.9)
Breastfeeding information from nurse/doula/midwife
No	24,475 (25.3)
Yes	76,333 (74.7)
Breastfeeding information from breastfeeding support groups
No	73,669 (75.9)
Yes	24,612 (24.1)
Breastfeeding information from family and/or friends
No	35,288 (35.3)
Yes	65,074 (64.7)
Breastfeeding information from breastfeeding hotline
No	86,345 (89.2)
Yes	11,366 (10.8)

n: Total sample size (unweighted), %: weighted percentages.

IPV, intimate partner violence.

Breastfeeding information and sociodemographics by IPV exposure

About 748 (67.9%) of prepregnancy IPV-exposed women and 663 (70.7%) of prenatal IPV-exposed women initiated breastfeeding. We found that only about 248 (21%) and 198 (23.1%) of those exposed to prepregnancy IPV and prenatal IPV, respectively, experienced early cessation. Over half the women exposed to prepregnancy IPV (58.8%) or prenatal IPV (60.8%) breastfed their baby for 6 months or more.

Approximately four in five prenatal IPV-exposed women (81%) were young adults, 3 in 10 (30.2%) were married, and over 3 in 5 (67.7%) had a household annual income of $28,000 or less. Less than two in five prepregnancy IPV-exposed women had a partner/husband who had more than a high school degree (38.6%) and was Black (23%) or Asian (6%), and over 3 in 4 (72.5%) lived in urban areas. One in four prenatal IPV-exposed women was Black (25.5%); approximately three in five were white (59.1%) and one in five was Hispanic (19.4%).

Approximately four in five women exposed to prepregnancy (79.7%) or prenatal (74.5%) IPV received breastfeeding information from their doctor. Over three in five women exposed to prepregnancy (68.6%) or prenatal (66.2%) IPV received breastfeeding information from their baby's doctor, and over three in five women exposed to pre-pregnancy (69.1%) or prenatal (67.3%) IPV received breastfeeding information from their nurse/doula/midwife. Over half the women exposed to prepregnancy IPV received breastfeeding information from family/friends (57.9%) and only 10.6% from breastfeeding hotlines. Table 2 summarizes the sociodemographic characteristics and breastfeeding behaviors of the sample population based on exposure to prepregnancy IPV or prenatal IPV.

Table 2.

Characteristics and Breastfeeding Behaviors of New Mothers in United States by Exposure to Intimate Partner Violence (2016–2018)

	Prepregnancy IPV		Prenatal IPV
	Yes, n (%)^a	No, n (%)	Yes, n (%)	No, n (%)
Breastfeeding behaviors
Breastfeeding initiation
No	312 (32.1)	12,188 (17.2)	259 (29.3)	12,241 (17.3)
Yes	748 (67.9)	57,441 (82.8)	663 (70.7)	57,526 (82.7)
Early breastfeeding cessation
Yes (≤1 week)	248 (21.0)	9,633 (24.4)	198 (23.1)	9,683 (24.3)
No (>1 week)	701 (79.0)	30,856 (75.6)	575 (76.9)	30,982 (75.7)
Breastfeeding duration
Short duration (<6 months)	745 (41.2)	30,490 (30.4)	596 (39.2)	30,639 (30.4)
Long duration (≥6 months)	1,064 (58.8)	69,959 (69.6)	925 (60.8)	70,098 (69.6)
Sociodemographic characteristics of participants
Woman's age
Teen mom (17–19 years)	161 (8.6)	4,654 (4.1)	125 (7.9)	4,690 (4.1)
Young adult mom (20–34 years)	1,584 (79.8)	79,683 (77.5)	1,334 (81.0)	79,933 (77.5)
Matured mom (≥35 years)	233 (11.7)	18,913 (18.4)	197 (11.1)	18,949 (18.4)
Woman's marital status
Married	571 (28.8)	62,438 (63.3)	475 (30.2)	62,534 (63.2)
Other	1,403 (71.2)	40,716 (36.7)	1,179 (69.8)	40,940 (36.8)
Woman's annual household income
Lower class (≤28,000)	1,336 (68.7)	38,591 (37.4)	1,132 (67.7)	38,795 (37.4)
Working class ($28,001–$40,000)	195 (11.5)	10,603 (10.7)	162 (13.4)	10,636 (10.6)
Middle class ($40,001–$100,000)	227 (13.0)	22,896 (24.2)	182 (13.3)	22,941 (24.2)
Upper class (≤$100,001)	98 (6.8)	22,853 (27.8)	79 (5.6)	22,872 (27.8)
Woman's highest education level
≤High school (<13 years)	1,001 (51.8)	37,491 (36.1)	830 (50.1)	37,662 (36.1)
>High school (13 years+)	960 (48.2)	64,724 (63.9)	811 (49.9)	64,873 (63.9)
Woman's residence
Urban	806 (72.5)	40,297 (77.3)	684 (75.1)	40,419 (77.3)
Rural	414 (27.5)	17,363 (22.7)	321 (24.9)	17,456 (22.7)
Woman's race
Black	539 (24.0)	18,757 (14.7)	487 (25.5)	18,809 (14.7)
Asian	59 (3.4)	7,002 (6.0)	56 (3.6)	7,005 (6.0)
White	852 (59.7)	59,466 (70.1)	691 (59.1)	59,627 (70.1)
Other	469 (12.9)	14,455 (9.2)	376 (11.8)	14,548 (9.2)
Woman's ethnicity
Hispanic	338 (18.2)	19,525 (18.9)	292 (19.4)	1,957 (18.9)
Non-Hispanic	1,600 (81.8)	80,609 (81.1)	1,333 (80.6)	80,876 (81.1)
Partner's characteristics
Partner's highest education level
≤High school (<13 years)	714 (61.4)	36,237 (40.1)	609 (60.9)	36,342 (40.1)
>High school (13 years+)	454 (38.6)	52,288 (59.9)	386 (39.1)	52,356 (59.9)
Partner's race
Black	291 (23.0)	13,788 (13.8)	274 (23.7)	13,805 (13.8)
Asian	50 (5.2)	5,923 (6.0)	43 (5.3)	5,930 (6.0)
White	581 (61.3)	56,577 (71.2)	479 (60.1)	56,679 (71.2)
Other	237 (10.5)	10,297 (9.0)	200 (10.9)	10,334 (9.0)
Sources of breastfeeding information
Breastfeeding information from woman's doctor
No	374 (20.3)	21,240 (22.3)	328 (25.5)	21,286 (22.2)
Yes	1,483 (79.7)	78,741 (77.7)	1,221 (74.5)	79,003 (77.8)
Breastfeeding information from baby's doctor
No	514 (31.4)	29,229 (31.1)	432 (33.8)	29,311 (31.1)
Yes	1,306 (68.6)	68,991 (68.9)	1,082 (66.2)	69,215 (68.9)
Breastfeeding information from nurse/doula/midwife
No	500 (30.9)	23,975 (25.3)	420 (32.7)	24,055 (25.2)
Yes	1,337 (69.1)	74,996 (74.7)	1,107 (67.3)	75,226 (74.8)
Breastfeeding information from breastfeeding support groups
No	1,314 (74.0)	72,355 (76.0)	1,079 (73.2)	72,590 (76.0)
Yes	493 (26.0)	24,119 (24.0)	428 (26.8)	24,184 (24.0)
Breastfeeding information from family and/or friends
No	727 (42.1)	34,561 (35.2)	608 (38.9)	34,680 (35.2)
Yes	1,102 (57.9)	63,972 (64.8)	917 (61.1)	64,157 (64.8)
Breastfeeding information from breastfeeding hotline
No	1,599 (89.4)	84,746 (89.1)	1,308 (88.4)	85,037 (89.2)
Yes	194 (10.6)	11,172 (10.9)	184 (11.6)	11,182 (10.8)

n: Population count (unweighted), %: weighted percentages.

IPV, intimate partner violence.

Association between IPV and breastfeeding behaviors

We found no significant association between women exposed to IPV in the 12 months before pregnancy or exposed to IPV during pregnancy, and breastfeeding initiation. We also found no statistically significant association between either type of IPV and early breastfeeding cessation. However, women exposed to IPV during pregnancy had statistically significantly lower odds (26%) of breastfeeding for 6 months or more (OR = 0.74, 95% CI = 0.58–0.94), while those exposed to prepregnancy IPV showed no significant association with breastfeeding for at least 6 months, compared to nonexposed women, as shown in Table 3.

Table 3.

Association Between Exposure to Intimate Partner Violence and Breastfeeding Behaviors in Women Living in the United States

	Breastfeeding initiation		Early breastfeeding cessation (≤1 week)		Breastfeeding duration (≥6 months)
	OR^a	95% CI	OR	95% CI	OR	95% CI
Prepregnancy IPV
No	1 (Ref.)		1 (Ref.)		1 (Ref.)
Yes	0.91	0.64–1.30	1.12	0.77–1.64	0.97	0.77–1.22
Prenatal IPV
No	1 (Ref.)		1 (Ref.)		1 (Ref.)
Yes	1.10	0.74–1.65	0.92	0.62–1.38	0.74	0.58–0.94
Woman's age
Teen mom (17–19 years)	1 (Ref.)		1 (Ref.)		1 (Ref.)
Young adult mom (20–34 years)	0.99	0.83–1.19	0.81	0.69–0.95	1.22	1.09–1.37
Matured mom (≥35 years)	0.85	0.69–1.04	0.75	0.62–0.91	1.27	1.11–1.44
Woman's marital status
Other	1 (Ref.)		1 (Ref.)		1 (Ref.)
Married	1.73	1.59–1.89	0.91	0.83–0.99	1.46	1.38–1.54
Woman's annual household income
Lower class (≤28,000)	1 (Ref.)		1 (Ref.)		1 (Ref.)
Working class ($28,001–$40,000)	1.62	1.44–1.82	0.92	0.82–1.04	1.60	1.48–1.73
Middle class ($40,001–$100,000)	1.85	1.67–2.07	0.75	0.67–0.84	1.11	1.04–1.20
Upper class (≤$100,001)	2.17	1.90–2.48	0.66	0.57–0.76	1.22	1.14–1.30
Woman's highest education level
≤High school (<13 years)	1 (Ref.)		1 (Ref.)		1 (Ref.)
>High school (13 years+)	2.03	1.85–2.22	0.79	0.72–0.87	1.15	1.09–1.22
Woman's residence
Rural	1 (Ref.)		1 (Ref.)		1 (Ref.)
Urban	1.26	1.16–1.37	0.94	0.86–1.02	0.96	0.91–1.01
Woman's race
Black	1 (Ref.)		1 (Ref.)		1 (Ref.)
Asian	1.39	1.03–1.88	1.09	0.80–1.48	1.03	0.92–1.15
White	1.07	0.89–1.28	0.99	0.82–1.19	1.17	0.98–1.40
Other	1.42	1.17–1.172	1.02	0.84–1.24	1.09	0.96–1.40
Woman's ethnicity
Non-Hispanic	1 (Ref.)		1 (Ref.)		1 (Ref.)
Hispanic	1.44	1.30–1.61	1.04	0.94–1.15	0.81	0.76–0.86
Partner's highest education level
≤High school (<13 years)	1 (Ref.)		1 (Ref.)		1 (Ref.)
>High school (13 years+)	1.84	1.68–2.02	0.80	0.72–0.87	1.33	1.26–1.41
Partner's race
Black	1 (Ref.)		1 (Ref.)		1 (Ref.)
Asian	1.28	0.95–1.74	0.86	0.62–1.17	1.07	0.90–1.29
White	1.76	1.48–2.10	0.81	0.68–0.96	1.11	0.99–1.24
Other	1.73	1.43–2.10	0.83	0.68–1.01	1.10	0.98–1.24

Please note that each column (breastfeeding initiation, early breastfeeding cessation, and breastfeeding duration) depicts a separate model.

Results that are statistically significant in our study (p < 0.05) are presented in bold.

All multivariate logistic models are adjusted by age, marital status, annual household income, highest education level, partner/husband's highest education level, residence, woman's race, partner's race, and woman's ethnicity; and the covariates: receiving information from woman's doctor, baby's doctor, nurse/doula/midwife, support group, family/friends, and breastfeeding hotline.

95% CI, 95% confidence interval; IPV, intimate partner violence; OR, odds ratio.

Modifying effects of breastfeeding information source on the association between IPV and breastfeeding behavior

Receiving breastfeeding information from the baby's doctor weakened the association between prenatal IPV and early cessation (OR = 0.37, 95% CI = 0.18–0.78) (p for interaction = 0.009). There was a stronger association between exposure to prenatal IPV and breastfeeding initiation among women who received breastfeeding information from their family/friends (OR = 2.46, 95% CI = 1.24–4.88) (p for interaction = 0.010) or from a breastfeeding support group (OR = 3.03, 95% CI = 1.17–7.88) (p for interaction = 0.023), respectively, compared to those who did not receive breastfeeding information from these sources.

The association between prenatal IPV exposure and breastfeeding duration was stronger in women who received breastfeeding information from their nurse/doula/midwife (OR = 1.76, 95% CI = 1.09–2.83) (p for interaction = 0.021), but weaker in women who used breastfeeding hotlines (OR = 0.50, 95% CI = 0.26–0.95) (p for interaction = 0.033). Receiving breastfeeding information from family/friends weakened the association between IPV before pregnancy and breastfeeding duration (OR = 0.67, 95% CI = 0.45–0.99) (p for interaction = 0.042) and receiving breastfeeding information from support groups strengthened the association between IPV before pregnancy and breastfeeding initiation (OR = 3.60, 95% CI = 1.52–8.55) (p for interaction = 0.004).

Discussion

Main findings

Prepregnancy IPV was not associated with any breastfeeding behavior. Prenatal IPV was associated with reduced breastfeeding duration, but was not significantly associated with breastfeeding initiation or early cessation.

Receiving breastfeeding information from family/friends modified the association between prepregnancy IPV exposure and breastfeeding duration, as well as the association between prenatal IPV and breastfeeding initiation. Receiving breastfeeding information from breastfeeding support groups modified the association between prepregnancy IPV or prenatal IPV and breastfeeding initiation. Receiving information from baby's doctor modified the association between prenatal IPV and early cessation. We also found that receiving breastfeeding information from the nurse/doula/midwife or from breastfeeding hotlines modified the association between prenatal IPV and breastfeeding duration.

Interpretation of findings

Prenatal IPV-exposed women's motivation to seek help²¹ and to protect their unborn/newborns anyway they can^16,17 may be undermined by prolonged exposure to IPV (prepregnancy) and a desensitization to violence. The negative association between prenatal IPV and breastfeeding duration in this study, while consistent with other studies,^3,22,23 may be explained by isolation from social networks, lack of spousal and family support, and/or restricted health care access commonly experienced by women exposed to IPV.^3,10 Although IPV has been found to be a risk factor against breastfeeding initiation in previous studies,^10,20 the lack of association in this study, while unexpected, is consistent with James et al.'s findings.²⁴

The advice of clinicians like pediatricians, nurses, and midwives is highly valued by pregnant women/new mothers and may influence their breastfeeding decisions.^15,25 This may explain why receiving breastfeeding information from the pediatrician weakened the association between prenatal IPV and early breastfeeding cessation in our study. It may also explain why breastfeeding information received from the nurse/doula/midwife increased the odds of breastfeeding for a duration of 6 months or more in prenatal IPV-exposed women. These findings highlight the importance of doctors and nurses in addressing breastfeeding barriers, and affecting policy changes²⁶ that will promote healthy breastfeeding behaviors among women in the United States.

It is important to note that receiving breastfeeding education or information from health care providers like doctors and nurses/doulas/midwives may need to be supplemented with additional breastfeeding services²⁷ to provide lasting improvements in breastfeeding behaviors on IPV-exposed women. Receiving breastfeeding information from breastfeeding support groups positively modified the effect of both types of IPV on breastfeeding initiation and may be attributed to the potential expansion of social support networks and options opened to pregnant women through support group activities. Interestingly, previous studies have shown that breastfeeding women may desire more interaction/discussion with mothers with first-hand experience of breastfeeding, more so than from their health providers with no experience of breastfeeding.²⁸

Family/friends with positive breastfeeding experiences can positively influence breastfeeding behaviors of IPV-exposed women,¹³ which might explain why receiving breastfeeding information from family/friends raised the odds of initiating breastfeeding in prenatal IPV-exposed women in our study. Negative experiences on the other hand may adversely affect breastfeeding behavior and may explain the reduced likelihood of breastfeeding for 6 months or more in prepregnancy-exposed women who receive breastfeeding information from family or friends. Irrespective of the direction of modification in this study, family and friends are strongly influential and should be targeted when creating interventions to promote healthy breastfeeding behaviors.

Inadequate breastfeeding skills and/or social support from family/friends reported by teenage mothers in previous studies²⁹ may explain why being a teenage mother in this study was a risk factor for early breastfeeding cessation compared to mature mothers. Similar to other studies,^24,30,31 education increased the odds of initiating breastfeeding, while rural dwelling was a risk factor against breastfeeding initiation³² and breastfeeding for long durations.³² These might be because educated women tend to have better access to breastfeeding “messages,” relevant policy changes, and national programs than uneducated women³³; and there is reduced breastfeeding awareness and knowledge available in rural areas.³³

Strengths and limitations

Using PRAMS's complex dataset accounts for underrepresentation/overrepresentation, increasing the generalizability of the study to the U.S. population of new mothers with live births. Despite this, there are several limitations to be considered with our findings. First, PRAMS data are cross-sectional and cannot be used to establish causality. Second, because the information on breastfeeding behavior as well as IPV experiences were self-reported, recall bias might be a limitation. Third, the association between prenatal IPV and duration may partly be from residual confounding.

Fourth, IPV prevalence may have been affected by an unwillingness to admit IPV exposure, resulting in underreporting. Although our study did not attempt to make this distinction, it is possible that some women who were exposed to prenatal IPV may have also been exposed to prepregnancy IPV, which may have impacted our results. It is important to note that these findings only apply to U.S. women with live births and not to women with fetal deaths, stillbirths, or abortions. Finally, this study did not assess the type of breastfeeding information received from family and friends (positive or negative information). It is possible that breastfeeding information received by the women in our study may have been positive or negative, medically accurate, or misinformation.

Recommendations

Further research to better understand underlying reasons for perceptions of inadequacy in health care providers to effectively discuss all aspects of breastfeeding is recommended. Although a woman's ‘intention to breastfeed’ can be a strong predictor of breastfeeding behavior³⁴ and might have impacted our results, it was not in the scope of our study. Further research to understand its impact on breastfeeding and IPV exposure is recommended. While our study did not attempt to distinguish between exposure solely to prepregnancy and exposure solely to prenatal IPV—due to the low prevalence of IPV in our study, further studies contrasting breastfeeding behaviors among women exposed only to prepregnancy IPV, only to prenatal IPV, and both prepregnancy IPV as well as prenatal IPV are recommended.

Although the use of breastfeeding hotlines as a source of breastfeeding information only modified the effect of prenatal IPV on breastfeeding duration, the coronavirus disease 2019 (COVID-19) pandemic has increased e-health utilization nationwide. This popularization of e-health resulting from lockdowns and social distancing may have increased the relevance of hotlines as a source of breastfeeding information.

We recommend further research using more recent data that will reflect the impact of COVID-19, to assess the modifying effect of hotlines as a source of breastfeeding information (our data date back to before the COVID-19 pandemic).

Conclusions

This is the first study—to our knowledge—that not only assesses the association of both prepregnancy and prenatal IPV on breastfeeding behavior nationally but also evaluates the modifying effect of the source(s) of breastfeeding information on the relationship between IPV and breastfeeding. The role of doctors, nurses/doulas/midwives, breastfeeding support groups, and family/friends in promoting healthy breastfeeding behaviors cannot be overemphasized. Despite the positive influence of doctors on breastfeeding behavior,^13,25 they may be underutilized in the fight to promote healthy breastfeeding behavior³⁵—especially among women who are exposed to IPV.

There is therefore a need for continuous evidence-based breastfeeding education/promotion interventions to improve breastfeeding knowledge among doctors and within communities; and to ensure more informed and accurate conversations about breastfeeding. We encourage the fostering of trusting relationships between doctors and their patients to facilitate appropriate information dissemination. Individualized interventions will allow for routine IPV screening, early identification of IPV-prone women, and provision of support and interventions to curb/eliminate IPV. It will also improve the accuracy of information used by new mothers when making breastfeeding decisions—hence improving pregnancy and breastfeeding outcomes.

Attitudes and beliefs of family/friends are influencing factors in decision-making for IPV-exposed women, and a mother's advice is highly cherished, sometimes even more highly than the advice of their doctor.¹⁵ Women who feel supported by family/friends are more confident, with an improved sense of self-worth (despite IPV exposure). Breastfeeding support groups can provide a sense of support and belonging, while providing a safe space to share informed ideas and discuss obstacles faced by breastfeeding mothers—including exposure to IPV—thereby promoting positive breastfeeding behaviors among new mothers.

Program interventions that engage both health providers and community members in the creation of a safe environment for pregnant women and new mothers will help promote confidential discussions about IPV experiences, available support, and possible next steps toward protecting IPV victims and their unborn child/newborn, while working with them on an individual level to foster healthy breastfeeding behaviors over time.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

This research received no external funding.

References

Centers for Disease Control and Prevention. The National Intimate Partner and Sexual Violence Survey: 2015 data brief—updated release. CDC Stacks, Public Health Publications. Available at https://stacks.cdc.gov/view/cdc/60893 (accessed June 24, 2020).

Alhusen

, Ray

, Sharps

, et al. Intimate partner violence during pregnancy: Maternal and neonatal outcomes. J Womens Health (Larchmt), 2015; 24:100–106.

Mezzavilla

, Ferreira

, Curioni

, et al. Intimate partner violence and breastfeeding practices: A systematic review of observational studies. J Pediatr (Rio J), 2018; 94:226–237.

Wong

, Chang

. Altered eating behaviors in female victims of intimate partner violence. J Interpers Violence, 2016; 31:3490–3505.

Moraes

, Tavares da Silva Tde

, Reichenheim

, et al. Physical violence between intimate partners during pregnancy and postpartum: A prediction model for use in primary health care facilities. Paediatr Perinat Epidemiol, 2011; 25:478–486.

Park

, Greene

, Melby

, et al. Postpartum depressive symptoms as a mediator between intimate partner violence during pregnancy and maternal-infant bonding in Japan. J Interpers Violence, 2019; 36:19–20.

Martin-de-las-Heras

, Velasco

, Luna-del-Castillo

, et al. Breastfeeding avoidance following psychological intimate partner violence during pregnancy: A cohort study and multivariate analysis. BJOG, 2018; 126:778–783.

World Health Organisation. Maternal, newborn, child and adolescent health Updated 2020. Available at https://www.who.int/maternal_child_adolescent/data/en/ (accessed December 11, 2020).

Obioha

, Martin

, Obioha

, et al. Association between skin-to-skin contact post-birth and breastfeeding behaviour: A cross-sectional study of Nigerian women using the 2018 Demographic Health Survey. J Glob Health Rep, 2021; 5:e2021101.

10.

Caleyachetty

, Uthman

, Bekele

, et al. Maternal exposure to intimate partner violence and breastfeeding practices in 51 low-income and middle-income countries: A population-based cross-sectional study. PLoS Med, 2019; 16:e1002921.

11.

Centers for Disease Control and

Prevention

. Breastfeeding report card: United

States 2020

. Updated

September 17

, 2020. https://www.cdc.gov/breastfeeding/data/reportcard.htm (accessed November 12, 2020).

12.

McCann

, Baydar

, Williams

. Breastfeeding attitudes and reported problems in a national sample of WIC participants. J Hum Lact, 2007; 23:314–324.

13.

Office of the Surgeon General (U.S.), Centers for Disease Control and Prevention (U.S.), Office on Women's Health (U.S.). The Surgeon General's Call to Action to Support Breastfeeding. Breastfeeding from the Public Health Perspective. Rockville, MD: National Center for Biotechnology Information, 2011.

14.

Radzyminski

, Callister

. Health professionals' attitudes and beliefs about breastfeeding. J Perinat Educ, 2015; 24:102–109.

15.

Office of the Surgeon General (U.S.), Centers for Disease Control and Prevention (U.S.), Office on Women's Health (U.S.). The Surgeon General's Call to Action to Support Breastfeeding. Barriers to Breastfeeding in the United States. Rockville, MD: Office of the Surgeon General (US), 2011.

16.

Bullock

, Libbus

, Sable

. Battering and breastfeeding in a WIC population. Can J Nurs Res, 2001; 32:43–56.

17.

Kendall-Tackett

. Violence against women and the perinatal period: The impact of lifetime violence and abuse on pregnancy, postpartum, and breastfeeding. Trauma Violence Abuse, 2007; 8:344–353.

18.

Backstrom

, Wahn

, Ekstrom

. Two sides of breastfeeding support: Experiences of women and midwives. Int Breastfeed J, 2010; 5:20.

19.

Renfrew

, McFadden

, Dykes

, et al. Addressing the learning deficit in breastfeeding: Strategies for change. Matern Child Nutr, 2006; 2:239–244.

20.

Misch

, Yount

. Intimate partner violence and breastfeeding in Africa. Matern Child Health J, 2014; 18:688–697.

21.

Zink

, Elder

, Jacobson

. How children affect the mother/victim's process in intimate partner violence. Arch Pediatr Adolesc Med, 2003; 157:587–592.

22.

Sipsma

, Magriples

, Divney

, et al. Breastfeeding behavior among adolescents: Initiation, duration, and exclusivity. J Adolesc Health, 2013; 53:394–400.

23.

Shroff

, Griffiths

, Suchindran

, et al. Does maternal autonomy influence feeding practices and infant growth in rural India?. Soc Sci Med, 2011; 73:447–455.

24.

James

, Taft

, Amir

, et al. Does intimate partner violence impact on women's initiation and duration of breastfeeding?. Breastfeed Rev, 2014; 22:11–19.

25.

Eidelman

. Breastfeeding and the use of human milk: An analysis of the American Academy of Pediatrics 2012 Breastfeeding Policy Statement. Breastfeed Med, 2012; 7:323–324.

26.

Sriraman

, Kellams

. Breastfeeding: What are the barriers? Why women struggle to achieve their goals. J Womens Health (Larchmt), 2016; 25:714–722.

27.

Feltner

, Weber

, Stuebe

, et al. Breastfeeding Programs and Policies, Breastfeeding Uptake, and Maternal Health Outcomes in Developed Countries Comparative Effectiveness Review, No 210. Rockville, MD: Agency for Healthcare Research and Quality (US), 2018.

28.

Hinson

, Skinner

, Lich

, et al. Factors that influence breastfeeding initiation among African American women. J Obstet Gynecol Neonatal Nurs, 2018; 47:290–300.

29.

Smith

, Coley

, Labbok

, et al. Early breastfeeding experiences of adolescent mothers: A qualitative prospective study. Int Breastfeed J, 2012; 7:13.

30.

Heck

, Braveman

, Cubbin

, et al. Socioeconomic status and breastfeeding initiation among California mothers. Public Health Rep, 2006; 121:51–59.

31.

Wiener

, Wiener

. Breastfeeding prevalence and distribution in the USA and Appalachia by rural and urban setting. Rural Remote Health, 2011; 11:1713.

32.

Senanayake

, O'Connor

, Ogbo

. National and rural-urban prevalence and determinants of early initiation of breastfeeding in India. BMC Public Health, 2019; 19:896.

33.

Acharya

, Khanal

. The effect of mother's educational status on early initiation of breastfeeding: Further analysis of three consecutive Nepal Demographic and Health Surveys. BMC Public Health, 2015; 15:1069.

34.

Linares

, Rayens

, Gomez

, et al. Intention to breastfeed as a predictor of initiation of exclusive breastfeeding in Hispanic women. J Immigr Minor Health, 2015; 17:1192–1198.

35.

Stolzer

, Hossain

. Breastfeeding education: A physician and patient assessment. Child Dev Res, 2014; 2014:1–6.