Associations Between Breastfeeding and Mother–Infant Relationships: A Systematic Review

Abstract

Introduction:

Breastfeeding has been associated to an improved mother–child bond, although this link is lacking the sufficient empirical support.

Aims:

The aims of this study were to carry out a systematic review as a continuation to the one by Jansen et al., and to clarify the link between breastfeeding and mother–infant relationships.

Materials and Methods:

Data sources: A search was conducted using PsycINFO, MEDLINE, and CINAHL (2008–2018), using both free text words and subject headings. Additional hand-searching was performed. Study selection: The Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) framework guided (the reporting of) the study selection and data extraction. Eligible articles were subsequently selected by title, abstract, and full text review. Data extraction: Data regarding study setting, participants, breastfeeding, mother–infant relationships, and outcome measures were extracted in a systematic way.

Results of Data Synthesis:

Ultimately, 13 articles were included. Most studies were conducted in Europe (46.15%) and were nonexperimental studies (92.3%). Eight studies were longitudinal, using between two and five assessment time points. The instruments used to assess the bond between mother and child showed great variability. From the studies, 8 (61.5%) employed self-report measures and four assessed maternal perception. Others used open-ended questions regarding the maternal lactation process or about the mother's activity while breastfeeding. Other measures used were maternal perception of her baby and personality variables associated to breastfeeding. Five articles studied the bond using external observations of the interactions between mother–child. The quality of the studies (Medical Education Research Studies Quality Instrument [MERSQI]) was overall fair to good.

Conclusion:

It is complicated to extract generalizable results because of the conceptual and instrumental variability of the mother–child relationship. This association is complex and the way in which breastfeeding is carried out would appear to be a decisive factor, influenced in turn by additional variables that should also be taken into account. The relationship indicators most frequently associated to breastfeeding are maternal sensitivity and secure attachment.

Introduction

Breastfeeding has many important benefits for both the baby's and the mother's health.^1,2 Certain underlying neurohormonal pathways have been found in relation to the positive effects of breastfeeding on the mother–child bond. In particular, it has been suggested that the hormones that regulate lactation (oxytocin and prolactin) have an effect on certain brain structures that enable maternal, protection and social behaviors, including attachment to the child, social memory, and recognition.³

Many pregnant women choose breastfeeding over bottle feeding because of the expected positive effects. This decision has important emotional repercussions (including feelings of guilt and emotional suffering in relation to their self-perception as mothers) on mothers who cannot breastfeed, despite their best efforts, because of external factors such as lack of milk production, mastitis, breast surgery, or other factors.^4,5

Nevertheless, the above-mentioned biological reasons have not been sufficient to empirically support the associations between breastfeeding and the mother–infant relationships. This could be because of several reasons, although mainly owing to the lack of empirical studies on the matter.⁶ The review by Jansen et al.⁶ considers the limitations associated to studying the above relationship, including the vast conceptual variability and the difference in instruments, methods, and information sources (mother, children) used in the analysis of these early mother–child relationships.

Since the previous review,⁶ to the best of our knowledge, in the past 10 years, no other reviews have been carried out, despite the interest that this issue has for the health care context. Moreover, the above-cited review⁶ was mostly focused on the endocrine and sensory factors involved in breastfeeding, and presented them as the mechanisms sustaining the associations between breastfeeding and early relationships.

This systematic review aims to assess the possible association between breastfeeding and the mother–child relationships. In addition, it has aimed to analyze the conceptual and methodological variability of these studies, and the possible psychological variables involved in this association.

Materials and Methods

This section is reported according to the Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) guidelines.⁷

Measures and definitions

This study focuses on the concept of the “mother–infant relationship (bond).” It is important to reiterate, although it has been mentioned previously, that there is a vast conceptual variability in relation to this concept. There are numerous terms that have been associated to it and that are used interchangeably. The mother–infant relationship consists of two complementary facets: the maternal bond or the tie from the mother to the infant, and infant attachment or the tie from the infant to the mother.

Infant attachment entails the tie between infant and mother that develops from a distinct innate behavioral system, promoting mother–infant proximity.⁸ The definition of the maternal bond (-ing) is less clear, although it could be defined as a biologically based emotional investment in the infant. It is characterized by the expression of a clearly defined set of maternal behaviors.⁹ In this context, “maternal sensitivity” can be defined as the mother's ability to accurately perceive the infant's signals and to respond to them appropriately and promptly.¹⁰

Data sources and search strategy

This systematic review synthesized research published in English between 2008 and 2018 regarding the associations between breastfeeding and the bond between mother and child. A search in PsycINFO (American Psychological Association), MEDLINE through PubMed (US National Library of Medicine), and CINAHL (EBSCOhost) was conducted. The search terms were selected to generate an overview of breastfeeding and the mother–child relationships. The terms used in combination were as follows: breastfeed* and bond*, breastfeed* and attachment, breastfeed* and mother–infant relation*, breastfeed* and early relation*. A snowball search strategy was also used to identify additional relevant articles from references in published research and reviews and to cross-check citations in these articles for any missed articles.

Inclusion and exclusion criteria

Empirical studies that presented a quantitative or qualitative association between breastfeeding and early relationships were eligible for inclusion. Specific inclusion criteria were as follows: (1) studies published in peer-reviewed journals, (2) used human participants, (3) included a measure of infant feeding (e.g., breastfeeding, formula feeding, bottle feeding, and breastfeeding duration), (4) included an indicator of the mother–infant relationship, such as direct measures of the relationship or indirect measures of maternal perception of either the relationship, the child's behavior or other variables associated to the relationship (i.e., maternal personality).

Exclusion criteria for the article were to not be available in English, studies with mothers or babies who had any diagnoses or other special circumstances, and studies that analyzed the association between the mother–child bond with breastfeeding from an exclusively physiological point of view.

Study selection

The search was performed by one author and was duplicated by a clinical librarian. The articles were reviewed by title and classified as clearly ineligible or potentially eligible. Two authors checked the first 100 titles, because agreement was high, a single researcher continued. Two researchers reviewed the abstracts of the potentially eligible articles, and finally, reviewed the full text of each remaining article. For articles that had no abstract, their full text was retrieved.

Data extraction and quality assessment

Data on study location, participants and setting, type of “mother–infant relationship” index used, type of breastfeeding measure, type of study, assessment time points and main findings were extracted by one researcher and duplicated by a second researcher, using a data extraction template. The Medical Education Research Studies Quality Instrument (MERSQI) was used to assess the methodological quality of the studies included,¹¹ because it has been shown to be a useful method for tracking and benchmarking the quality of medical education scholarship in obstetrics and gynecology.¹² As an additional quality measure, whether studies had been adjusted for other possible confounders was recorded.

Data synthesis and analysis

The largest difficulty for data synthesis and analysis was in relation to the wide variety of indicators, both direct and indirect, that are used as measures of the “mother–infant relationship (bond).” To be able to reduce and organize that wide variety of terms and concepts, a descriptive qualitative analysis was conducted, assisted by ATLAS.ti qualitative research software. The full text articles were entered into ATLAS.ti and coded as primary documents, creating a hermeneutic unit.¹³ The analysis of the information in these registries was carried out using internal procedures associated to the “Grounded Theory.”¹⁴ Initially, different quotations were identified so as to later create codes. This approach was developed to systematically and critically analyze diverse sources while minimizing inaccuracy and bias.¹⁵

Four codes were created: “mother's self-perception regarding bond with child” (including the “self-report measures” and “maternal sensitivity” quotations), “external observation of behaviors associated to the mother, baby or their interactions” (including the “external observation” and “behavior” quotations), “mother's perception of child's behavior” (including the “mother's perception” and “child's behavior” quotations), and “variables associated to maternal personality” (including the “maternal personality” quotation). The above-mentioned quotations had to exist simultaneously throughout each article; this was a requisite for them to be included in the registry. The analysis of the texts for revision ended once theoretical saturation was reached, that is to say, once the information became repetitive.

This process was carried out, in parallel, by two researchers, one of them belonging to the team and the other was external. After the data collation (quotations and codes), it was found that both had generated the same categories.

Results

Search and study characteristics

The search yielded 711 unique hits. After title, abstract, and full text review, 13 articles were included in this review. Figure 1 provides the flow of the selection process. Most studies were conducted in Europe (46.15%) and were nonexperimental studies (92.3%). Eight studies were longitudinal, using between two and five assessment time points. Together, the studies included assessments time points ranging from the first trimester of pregnancy to 6 years after childbirth. Four studies (30.76%) used correlational methods, at 48–72 hours postpartum, 3 months after childbirth, and 5–6 months after childbirth.

FIG. 1.

Flowchart of the selection process.

Breastfeeding was mostly assessed (84.62%) by means of self-report questions regarding the type of lactation and duration. Only two studies used specific scales, specifically, the Breastfeeding Assessment Score Tool (LATCH)¹⁸ and the Infant Breastfeeding Assessment Tool (IBFAT).⁴⁴

The instruments used to assess the bond between mother and child showed great variability. From the studies, 8 (61.5%) used self-report measures, 4 assessed maternal perception,^16,18,38,41 with 2 using the Maternal Attachment Inventory (MAI).^16,18 The remaining two used open-ended questions regarding the maternal lactation process³⁸ and the mother's activity while breastfeeding.⁴¹ One article analyzed the perception of the mother about her baby.³⁹ Three of the articles assessed personality traits of the mother associated to breastfeeding.^23,26,44 Five of the 13 articles (38.5%) studied the bond using external observations of the interactions between them.^{20,34,47,49,50}

The quality of the studies was overall fair to good (Table 1). Specifically, 12 articles (quantitative research) could be scored in all six of the MERSQI domains. The average MERSQI score was 12.37 (±2.44) with scores ranging from 8 to 16 (median 13). No statistically significant differences were found based on the coding. Six studies did not address confounding variables.

Table 1.

Study Characteristics, Findings, and Quality of the Studies

Author (year), country [number of institutions]	Type of study: Time points assessment	MERSQI score ^a	Sample Special included/excluded criteria	Assessment, instruments, and dimensions	Result(s) regarding relation between breastfeeding and mother–infant relationship
Abuhammad (2016),¹⁶ Jordan [3]	Quasiexperimental: 2 and 12 months after childbirth	11	258 mothers. Three groups (breastfeeding, formula milk, mixed feeding)	Mother's self-report: MAI¹⁷ Dimension: maternal sensitivity. Ad hoc self-report for infant temperament measure (six items). Ad hoc self-report for breastfeeding (categorical: breast, formula milk, mixed feeding)	Higher maternal sensitivity for breastfeeding versus formula feeding mothers at 2 and 12 months (p < 0.001), controlled for infant temperament. Higher maternal sensitivity for formula feeding versus mixed method mothers at 2 months (p < 0.001).
Cetisli et al. (2018),¹⁸ Turkey [1]	Correlational: 48–72 hours after childbirth	8	175 first time mothers	Mother's self-report: MAI. Dimension: maternal sensitivity. LATCH Breastfeeding Assessment Score Tool¹⁹ Dimensions: audible swallowing, type of mother's nipple, comfort of the mother, mother's needs to hold the baby against the breast	Positive relation between the MAI score and LATCH total (p < 0.000). Other results: MAI (p = 0.003) and LATCH score (p = 0.015) was higher in mothers who had vaginal birth than in mothers who had cesarean birth
Gibbs et al. (2018),²⁰ United States [1]	Longitudinal: (T1) 9 months after childbirth. (T2) 2 years after childbirth	14	8900 mothers and their children	(T1): Mother's self-report: Ad hoc self-report for breastfeeding: type of feeding and duration (months). (T2): Observational assessment (home visits by researcher). Infant attachment behavior (TAS-45²¹). Dimensions: infant attachment security and temperamental dependence. Baby behavior²² Confounding factors: Parenting dimensions: frequency of reading to the child, maternal cognitive stimulation and maternal sensitivity, and maternal warmth household characteristics (family socioeconomic status, family structure, maternal education). Child characteristics (weight, race and ethnicity, twin, gestation age, sex). Mother characteristics (age, nationality, body mass index, smoking and maternal depression).	Mothers who predominantly breastfeed have higher socioeconomic status and education, lowest rate of children attending daycare. Mothers who predominantly breastfed have the highest scores for reading (to the child) frequency (T1 and T2), maternal cognitive stimulation, maternal sensitivity, and maternal warmth. Children predominantly breastfed 6 or more months score higher on infant attachment security than children not predominantly breastfed (p < 0.001, modest associations). Adding parenting measures, other infant feeding practices, and household, child and maternal characteristics reduced the association. Other parenting behaviors such as reading to the child (β = 0.04, p < 0.001) or maternal warmth (β = 0.04, p < 0.001) have modest associations with infant attachment security. Breastfeeding and other parenting measures are not associated with temperamental dependency. Children predominantly breastfed were rated as more “warm and cuddly,” “cooperative,” and were less likely to be considered “demanding and angry” compared with 2-year-olds who were predominantly breastfed for <3 months
Keller et al. (2016),²³ Croatia [1]	Correlational: 6 months after childbirth	12	303 mothers who had stopped breastfeeding and with a child of up to 2 years of age	Mother's self-report: Thompson's International English Big-Five Mini-Marker²⁴ Dimensions: extraversion, openness, neuroticism, conscientiousness and agreeableness. Postpartum Bonding Questionnaire (PBQ)²⁵ Dimensions: “impaired bonding,” “rejection and anger,” “anxiety about care,” and “risk of abuse.” Ad hoc self-report for breastfeeding (duration in months)	Mother's “agreeableness,” “openness,” and “age” strongly and positively correlate with breastfeeding duration. Weak but significant negative correlation (p = 0.04) was found between “neuroticism” and breastfeeding duration. “Anxiety about care” decreasing the likelihood for breastfeeding beyond minimal 6-month recommended period (p = 0.009).
Kestler-Peleg et al. (2015),²⁶ Israel [8]	Longitudinal: (T1) Third trimester of pregnancy. (T2) 6–8 weeks after birth. (T3) 4–5 months after birth	13.5	236 mothers	Mother's self-report: Ad hoc Breastfeeding Motivation Scale (BMS) based on the SDT²⁷, (T2 and T3). Dimensions: intrinsic, integrative and identified regulations constituting autonomous motivations, and introjected and two external regulations constituting the controlled motivations. Positive and Negative Affect Schedule (PANAS)²⁸ (T2 and T3). Edinburgh Postnatal Depression Scale (EPDS)²⁹ (T2 and T3). Satisfaction with Life Scale (SWLS)³⁰ (T2 and T3). Maternal Self-Efficacy Scale based on Parenting Sense of Competence (PSOC) questionnaire³¹ and in additional relevant statements³² (T2 and T3). Experiences in Close Relationship Scale (ECR)³³ (T1). Dimensions: avoidant and anxious attachment. Sociodemographic variables (T1), gestational week, number/weight of babies born (T2). Ad hoc self-report for breastfeeding (T2 and T3) (yes/no).	Motivation to breastfeed to enhance the baby's health that stems from external regulation decreased over time (p < 0.05), with small effect size (0.02). The other five breastfeeding motivation subscales remained stable over time (p > 0.05). Although SDT was not associated with actual breastfeeding behavior, it was linked to maternal state of mind. Autonomous breastfeeding motivations were positively related to positive aspects of maternal well-being (positive effect, life satisfaction) and self-efficacy. Controlled breastfeeding motivations were positively related to negative aspects of well-being (negative effect, postpartum depression) and negatively related to maternal self-efficacy. Identified regulation (to feel like a “good mother”), which reflects a relatively autonomous breastfeeding motivation, was positively related to negative aspects of well-being (negative effect, postpartum depression). Women who breastfeed out of extrinsic motivations, even if they are autonomous motivations, may experience more negative emotions and symptoms of postpartum depression, and fewer positive emotions; they may also feel less competent. Avoidant attachment did not predict breastfeeding motivation at either time. Anxious attachment predicted identified regulation (to feel like a “good mother”), introjected regulation (significant pressure by others) and external regulation (instrumental needs). Breastfeeding motivation seems to be affected by anxious attachment irrespective of the level of autonomy attached to the motivation. Other results: Nonsignificant correlations among breastfeeding motivation dimensions and sociodemographic and gestational variables
Kielbratowska et al. (2015),³⁴ Poland [1]	Correlational: 3–5 months after birth		98 mothers	Observational Scale of the Children Development Scale (DSR)³⁵ Observational measure, by a psychologist. Dimensions: vigor, adaptability, regularity (mother's answers) and sensitivity (mother's answers).Mother and Baby Scale^36,37 Observational measure, by mother. Dimensions: alertness during feeds, irritable during feeds, lack of confidence in feeding, easiness and global confidence.Ad hoc self-report for breastfeeding (breastfeeding versus bottle feeding).	Breastfed infants scored higher on vigor (activity and initial reaction dimensions) (p = 0.033) than bottle-fed infants. Bottle-fed children scored higher on regularity than breastfed children (p = 0.005).Mothers who bottle fed perceived themselves to be less confident in feeding (p < 0.001) and perceived their babies to be more irritated during feeding (p = 0.064) than mothers who breastfed
Kronborg et al. (2015),³⁸ Denmark	Longitudinal:(T1) 1–2 weeks after childbirth.(T2) 6 months after childbirth.Qualitative study at T2	No applicable	108 first time mothers who initiated breastfeeding at T1	Open ended questions regarding the lactation period.	Three overlapping phases were identified:First phase: on shaky ground, characterized by breastfeeding interwoven with mothering, painful breastfeeding, and conflicting advice.Second phase: searching for a foothold, characterized by reading the baby's cues, concerns about milk production, for or against breastfeeding, and looking for professional support.In this second phase, regardless of the mother's choice of feeding method, the positive relationship with the newborn is highlighted. The relationship with the breastfed child was often described through a focus on the difficulties in understanding the baby's cues, overcoming the difficulties regarding lactation (especially pain) and being self-efficient enough in taking care of the child.Regardless of their choice, the mothers described joy, confidence, and the loving smile of the child when being fed.Health professional support seems to be a determining factor in the breastfeeding process. Their acceptance of the mother's preference benefits wellbeing and decreases feelings of guilt and suffering.Third phase: at ease with choice of feeding, characterized by a thriving baby, trust in breastfeeding capability, and approval of feeding preference.Together, these themes and subthemes constituted the overall theme: being on a breastfeeding–bonding trajectory.
de Lauzon-Guillain et al. (2012),³⁹ United Kingdom [1]	Correlational:3 months after childbirth	11	316 mothers	Mother's self-report:Revised Infant Behavior Questionnaire.⁴⁰ Dimensions: extraversion, negative affectivity, orienting/regulation.Ad hoc self-report for breastfeeding (breast, formula milk, mixed feeding).	Breastfed and mixed-fed infants had lower scores in surgency/extraversion (smiling and laughter, high pleasure, vocal reactivity) (p < 0.001) and orienting/regulation (soothability, reduction of fussing, crying, or distress when caregiver uses soothing techniques) (p = 0.01) compared with formula-fed infants.Breastfed infants and mixed-fed infants had higher scores in negative affectivity (falling reactivity) compared with formula-fed infants (p = 0.03).Breastfed infants had lower scores for approach (extraversion) and low pleasure (orienting/regulation) and higher scores for distress to limitation (negative affectivity) compared with formula-fed infants
Liu et al. (2013),⁴¹ China [4]	Longitudinal (two measures in relation to children's age):(T1) starting preschool (T2) 6 years old	14	1267 mothers	(T1) when children were starting preschool (retrospective):Mother's self-report:Ad hoc self-report for breastfeeding: categorical (breastfed, formula, mixed) and duration (<7 months, between 7 and 10 months, >10 months).Ad hoc maternal interaction measure (“Did you talk to the child while (breast) feeding in the first 2 years.” Active bonding (yes/no).Parents' self-report:Social adversity: 11 variables⁴²(T2) when children were 6 years old.Internalizing Behavior scale from the Child Behavior Checklist (CBCL)⁴³ Dimensions: emotionally reactive, anxious/depressed, withdrawn and somatic complaints.	Children who were exclusively breastfed had significantly lower mean scores for somatic complaints (p = 0.03) and total internalizing problems (p = 0.04).Active bonded children had significantly lower scores across emotionally reactive (p = 0.03), anxious/depressed (p = 0.02), withdrawn syndromes and total internalizing problems (p = 0.001).Exclusive breastfeeding with active bonding displaying the lowest scores in internalizing problems, followed by nonexclusive breastfeeding with active bonding. Nonexclusive breastfeeding without active bonding had the highest scores, after controlling for adversity (p ≤ 0.001) and gender (p ≥ 0.212).Main effect for breastfeeding duration on anxious/depression (p = 0.038), somatic complains (p = 0.039), and total internalizing (p = 0.051) were found. The longer the duration the less internalizing behavior.Other results:Social adversity scores were significantly lower for those whom actively bonded with their baby while feeding (p < 0.001), indicating they had a better socioeconomic background and health status.Social adversity was positively correlated to each syndrome and total internalizing problems (p < 0.001).
Loke and Chan (2013),⁴⁴ China [1]	Longitudinal:(T1): 2–4 hours after childbirth.(T2): 24 hours after childbirth.(T3): 6 weeks after childbirth	13	202 mothers and their children (delivered vaginally)	(T1 and T2). Assessment by midwives.Infant Breastfeeding Assessment Tool (IBFAT)⁴⁵ Dimensions: infant's alert state, infant's readiness to feed or arousability, rooting, fixing and sucking pattern, mother's satisfaction with the feeding.(T3) Mother's self-report by telephone.Breastfeeding Self-Efficacy Scale–Short Form (MBSES-SF).⁴⁶Ad hoc self-report for breastfeeding (exclusive breastfeeding, complementary breastfeeding, and bottle feeding).	Women who exclusively breastfed their infants at 6 weeks had higher levels of breastfeeding self-efficacy than those who complementary breastfed and bottled fed their infants (p < 0.001).Characteristics and breastfeeding behaviors of newborns at T1: Newborns who were being exclusively breastfed at 6 weeks had a higher readiness to feed, rooting, fixing, and a stronger sucking pattern than those who were not. There was a significant association between the IBFAT score and the breastfeeding outcome (p = 0.023).Significant contributing factors for exclusive breastfeeding at T3 were women who planned to breastfeed for at least 6 months (p ≤ 0.001), women with a high level of breastfeeding self-efficacy (p ≤ 0.001), and women whose infants had high IBFAT scores (p = 0.045)
Papp (2014),⁴⁷ United States [31]	Longitudinal:1, 6, 15, 24, and 36 months after childbirth	16	1306 mothers and their children	Demographic and early characteristics at 1 and 6 months: Maternal race and ethnicity, age, education, and marital or cohabiting status at 1 month old. Home Observation of Measurement of the Environment (HOME⁴⁸), maternal depressive symptoms, and an observer rating of the overall impression of the mother–infant relationship.Breastfeeding assessed at 1, 6, 15, 24, and 36 months (yes/no).Mother–child interaction qualities. Semistructured by trained observers 15-minute mother–child interactions in the home at 6 and 15 months and in the laboratory at 24 and 36 months. Dimensions: child positivity, child negativity, and maternal sensitivity.	Breastfeeding was positively associated with maternal sensitivity (p < 0.001) and child positivity (p = 0.022) over time, but was not reliably associated with changes in child negativity (p = 0.188).Current breastfeeding was associated with increases in maternal sensitivity over time, demographic controls and early characteristics of total HOME score, maternal depression, and observed mother–child relationship quality.Once the demographic controls and early characteristics were accounted for, breastfeeding was not reliably associated with trajectories of observed child behavior (i.e., negativity, positivity) during mother–child interaction across early childhood.
Smith and Ellwood (2011),⁴⁹ Australia [1]	Longitudinal:3, 6 and 9 months (registering for 7 days at each time point)	9	188 mothers	Mother's registry:Activities, including feeding, over a 7-day period by pressing coded buttons on an electronic TimeCorder device. Dimensions (time spent): (1) feeding and (2) emotional care (other interactive care activities including soothing, holding or cuddling the infant).Mother's self-report:Ad hoc self-report for breastfeeding (exclusively breastfeeding at 3/6 months, breastfeeding and feeding solids at 6/9 months, exclusive formula feeding at 3/6 months, formula feeding and feeding solids at 6/9 months).	Feeding time and frequency depend on exclusivity of breastfeeding more than on the age of the infant. Exclusivity of breastfeeding takes more time.Total time and frequency of emotional care seems to be a function of exclusivity of breastfeeding rather than of age. High emotional care in exclusivity of breastfeeding.Maternal time spent in emotional care correlates strongly with time spent on breastfeeding. Infants elicit frequent episodes of emotional care from their mothers. This also shows a pattern of high time use, and reduced maternal time commitment once the infant has ended the exclusive breastfeeding phase. Breastfeeding infants receiving solids at 6 or 9 months received less emotional care time than exclusively breastfed infants, but more than nonbreastfed infants.The evident synchrony in the timing of breastfeeding and emotional care during the day suggests a relationship between these activities—although the causal direction is not clear.The less exclusive the breastfeeding, and the older the infant, the less time spent on childcare
Tharner et al. (2012),⁵⁰ the Netherlands [1]	Longitudinal:30 weeks pregnant, and 2, 6 and 14 months after childbirth	14	628 mothers and their children	Maternal Sensitive Responsiveness (mother's observation; laboratory)¹⁰ At 14 months after birth.Attachment Security and Disorganization (children's observation; laboratory). Strange Situation Procedure (SSP)¹⁰ Dimensions: attachment security, disorganization. Attachment types: secure, insecure–avoidant, insecure–resistant.OXTR Genotype (mothers).Background variables: mother (age, parity, educational and socioeconomic level), birth weight, gender of the child, 5-minute APGAR score, and type of delivery.Family functioning: Family Assessment Device (maternal self-report at 30 weeks of pregnancy).Maternal psychological well-being: Brief Symptom Inventory (maternal self-report at 2 months postpartum).Ad hoc self-report for breastfeeding. 24–48 hours postpartum and 2 and 6 months after birth (categorical: yes/no and duration of breastfeeding).	Longer duration of breastfeeding (for at least 6 months) was related to more maternal sensitive responsiveness (p = 0.02), more attachment security (p = 0.03), and less attachment disorganization (p = 0.02), adjusted for child gender, type of delivery, birth weight, 5-minute APGAR score, maternal age, prenatal family functioning, and maternal postnatal depressive symptoms.No significant association between duration of breastfeeding and the different categories of infant–mother attachment, but longer duration of breastfeeding predicted a lower risk of disorganized attachment classification (p = 0.04).Children who were exclusively breastfed at 2 months had significantly lower disorganization scores than children who had never been breastfed (p = 0.05).

Maximum MERSQI score is 18.

MAI, Maternal Attachment Inventory; MERSQI, Medical Education Research Studies Quality Instrument; SDT, Self-Determination Theory.

Mother's self-perception regarding bond with child

Four articles^16,18,38,41 used mother's self-report about attachment, with two using the MAI.^16,18 The results showed the importance of self-perception of maternal sensitivity on breastfeeding, both in relation to commencing (in first time mothers)¹⁸ and maintenance of the behavior.¹⁶ It was also found that vaginal birth influenced both maternal sensitivity and breastfeeding.¹⁸ Maternal behaviors aimed toward the baby, such as talking to them while feeding (active bonding), as part of maternal sensitivity, appeared to associate with a positive emotional relationship and to lower levels of somatization (when children were 6 years old).⁴¹

Breastfeeding with active bonding was shown to be, in this sense, the healthiest option, followed by nonexclusive breastfeeding with active bonding, thus, highlighting its importance.⁴¹ Finally, the qualitative review³⁸ showed that feeding created positive self-perception and emotions among first time mothers, regardless of the method used, and emphasized that the support by health professionals was a determining factor in the mother's wellbeing.

External observation of behaviors associated to the mother, baby, or their interactions

Five articles^{20,34,47,49,50} assessed the bond between mother and child by means of external observation of the behavior of the mother, the child, or their interactions. One of them²⁰ showed that the mothers of babies who were primarily breastfed showed higher scores of frequency of reading to the child, maternal cognitive stimulation, maternal sensitivity, and maternal warmth. Likewise, children predominantly breastfed 6 or more months scored higher on infant attachment security than children not predominantly breastfed, finding other covariables such as mother, infant, or domestic characteristics that could reduce the strength of this association.^20,50 Longer duration of breastfeeding predicted a lower risk of disorganized attachment classification.⁵⁰

Another study found that breastfeeding was positively associated with maternal sensitivity and child positivity over time, although the association between breastfeeding and child positivity disappeared if demographic and early characteristics were controlled for.⁴⁷ It was also found that breastfed infants scored higher on vigor and lower on regularity than bottle-fed infants.³⁴ Finally, it would appear that the total time and frequency of emotional care were a function of exclusivity of breastfeeding rather than of age.⁴⁹

Mother's perception of child's behavior

One article³⁹ assessed the mother's perception of her baby, using the Revised Infant Behavior Questionnaire (self-report). The results showed that breastfed and mixed-fed infants were perceived by their mothers as having lower scores on extraversion (tendency to associate with others and be active) and higher scores in negative affect (aversive emotional states such as irritability or fear) compared with formula-fed infants. Neonates seemed to experience the initiation of breastfeeding as more stressful than the initiation of formula feeding.

Variables associated to maternal personality

Three of the articles^23,26,44 analyzed the role of maternal personality as an indicator of the mother–child bond. Mother's agreeableness and mother's openness favored the continuity of breastfeeding, whereas mother's neuroticism and anxiety about care reduced the likelihood of breastfeeding after 6 months of age.²³

The mother's motivation to breastfeed was shown to be a fundamental variable in their well-being. The different motivations underlying breastfeeding associated maternal well-being and self-efficacy in different ways. On the one hand, when the motivation to breastfeed was autonomous (when it is engaged in for itself, without any interest in or expectation of action-related outcomes), the association to maternal well-being and self-efficacy was found to be positive; on the other hand, controlled breastfeeding (when it is motivated by interests other than the activity itself, such as to lose weight more quickly or to save on formula costs), was associated negatively.

Identified regulation was related to negative aspects of well-being.²⁶ Identified regulation involves a cognitive recognition of the underlying value of breastfeeding; it is perceived as important for the self (i.e., to feel like “a good mother”) even if it is not necessarily interesting or pleasurable. Finally, regarding self-efficacy, women who exclusively breastfed their infants at 6 weeks had higher levels of breastfeeding self-efficacy than those who complementarily breastfed and bottled fed their infants.⁴⁴

Discussion

The literature review carried out in this study regarding the associations between breastfeeding and early mother–child relationships has brought to light its complexities and the difficulty in establishing general conclusions, mainly because of the methodological variability. In particular, it is of interest to highlight the complexity of measuring “early mother–child relationships,” as several different conceptual variables are involved. The relationship can be assessed using two types of measures, on the one hand, those that directly measure the interactions, mainly through observational methods, and on the other hand, those that are indirect, using either maternal personality or maternal perception in relation to the bond or the child's behavior. In relation to self-report measures, although there is no consensus, the Postpartum Bonding Questionnaire (PBQ)⁵¹ and the MAI¹⁷ have been recommended.

One of the interesting aspects of this review is that only 5 of the 13 studies selected used a direct measure of the mother–child interactions, by means of external observation.^{20,34,47,49,50} In general, the results show that the time invested in breastfeeding or its quality in relational terms seem to be differential explanatory factors in the positive relationship between breastfeeding and mother–child interactions.

Specifically within the mother–child interaction, some studies^{20,34,39,44,47,50} have focused on the child's behavior. In these studies, two types of measures have been used: the child's behavior in itself and attachment, both being assessed either by the mother or by an external observer. Results have found a relationship between the child's behavioral characteristics and breastfeeding, independently of the type of measure used. Breastfeeding was associated with more adaptive behavioral responses³⁴ such as being “warm and cuddly” and “cooperative.”^20,47

Nevertheless, it should also be pointed out that the inclusion of other additional variables, such as parenting measures, household and child and maternal characteristics reduced the association²⁰ or even eliminated it completely.⁴⁷ This could be indicating the influence of additional variables, external to breastfeeding, in the early mother–child relationships. With regard to attachment,^16,18,20,47 longitudinal studies have found associations between secure attachment and breastfeeding.^20,50 Additional variables that have been found to influence this relationship are breastfeeding duration, parenting behaviors, maternal warmth, household, and child and maternal characteristics.²⁰

An especially relevant issue is the role that different personality variables play in the study of the association between breastfeeding and the mother–infant relationships. Thus, although some studies consider that certain maternal personality variables are a direct indicator of the bond,^{16,18,23,26,38,41,44} others^20,50 have included them as variables to control for because of their influence on the bond (i.e., maternal cognitive stimulation, warmth, depression, or psychological well-being). Among the first type, the main measure for the bond is “maternal sensitivity.”^{16,18,34,47,50} In fact, of the above-mentioned studies that used direct (observational) measures of the mother–child interactions, three of them assess maternal sensitivity solely based on the observation of the mother–child relationship.^34,47,50 The rest of the studies^16,18 measure maternal sensitivity using self-reports, specifically the MAI. Correlational and longitudinal studies that used observational assessments^34,47,50 suggest that mothers who predominantly breastfeed have the highest scores for maternal sensitivity, and that this association remains at 6 months,⁵⁰ 1 year,¹⁶ and 3 years of age,⁴⁷ even after controlling for child temperament. These studies suggest mediating variables such as vaginal birth,¹⁸ sociodemographic variables, and early characteristics of the environment and maternal depression.⁴⁷

In addition to maternal sensitivity, other characteristics of lactating mothers, versus alternative types of feeding, have been analyzed as the reason underlying the associations between breastfeeding and early maternal relations. Maternal breastfeeding self-efficacy has been associated positively and significantly with breastfeeding,^23,34,44 both in correlational^23,34 and longitudinal studies up to 6 weeks after childbirth,⁴⁴ and both with direct^34,44 and inverted measures (anxiety about care).²³

The real motivations of the mother to breastfeed and their origin seem to play a fundamental role in the associations between breastfeeding and mother–child relationships. Women who breastfeed out of extrinsic motivations, even if they have autonomous motivations (i.e., identified regulation), may experience more negative emotions and symptoms of postpartum depression, and fewer positive emotions; they may also feel less competent. Likewise, anxious attachment predicted identified regulation (to feel like a “good mother”), introjected regulation (pressure from significant others), and external regulation (instrumental needs such as to lose weight more quickly or to save on formula costs).²⁶ Finally, certain personality characteristics of the mother are associated to making it easier to establish interpersonal relationships (agreeableness and openness) and have also been associated positively to breastfeeding duration.²³

Limitations

This systematic review has several limitations. The main limitation has been the difficulty in determining a specific definition for mother–infant relationships (bond). If bond was conceptualized as the tie between mother and child that promotes maternal behaviors aimed toward proximity and care,⁶ its assessment should be carried out from pregnancy up to the first year of life of the baby, with repeated measures throughout the process. In most of the articles the gestational period was not taken into account and in some cases single time measures were used, therefore making it impossible to assure a cause–effect relationship between breastfeeding and the mother–child bond.

Another limitation is the diversity of assessment instruments used. Given the lack of a validated instrument to measure the bond between mother and child, it has been assessed using self-report measures or observation of the mother–child behavior, coded on scales by researchers. These measures assess different dimensions of the maternal bond, but not the construct in itself. In addition, it seems that the development of the bond though breastfeeding can be affected by different situations, family variables, maternal characteristics, and even by health professionals. All these factors complicate even further the possibility of determining if breastfeeding really does favor maternal bonding.

The fact that neither mothers nor babies with diagnoses or special circumstances were included can also be mentioned as a limitation, as their study could shed light on more specific and sensitive indicators. Nevertheless, their inclusion would have increased heterogeneity even further and would have therefore posed another major difficulty in establishing generalizable conclusions. It would thus be of great interest to subject this matter to future research. Finally, another limitation found was the lack of studies researching this topic; this limitation is reflected in the small number of articles contained in this review.

Implications for clinical care

Some of the results found suggest that the manner in which a mother breastfeeds has a decisive influence upon the positive effects of breastfeeding on the mother–child relationship. Furthermore, variables such as autonomous motivations and maternal breastfeeding self-efficacy have been suggested to be modulating variables of these effects. These results, from our point of view, have practical implications in the assistance of pregnant women by health professionals, specifically midwives.

In the first place, in some countries or contexts in which mothers feel pressured to breastfeed,^5,52 midwives should be trained to assess maternal preferences regarding lactation. They should also assess mothers about establishing a bond with their babies through breastfeeding, taking into account the obstacles they may come across and working on them. In this context, the numerous validated and well-known interventions aimed at assessing and increasing breastfeeding self-efficacy and to reducing anxiety regarding caring for the baby are important tools that should be included in the health care of pregnant mothers.^23,34,44

Another of the essential competencies of midwives within health care systems is to teach mothers how to identify and interpret signs from the baby and how to respond appropriately. On numerous occasions mothers express worries regarding breastfeeding as, in comparison with bottle feeding, they do not have feedback regarding the amount of milk that the child has obtained and they do not know if it is enough.³⁸ Helping mothers to identify and to manage the cues from the baby and to detect nonadaptive mother–child perceptions would improve not just breastfeeding, but it would also provide more peace of mind and wellbeing to mothers, which in turn, would improve the bond.

In the third place, it would be of the utmost importance to make mothers aware of the value of the time dedicated to their children in relation to the bond, regardless of the type of feeding used. Talking to the baby while breastfeeding or active bonding have shown to be the fundamental keys to the benefits of breastfeeding, compared with artificial lactation, in the emotional management of the newborn.^20,41 Finally, it should be highlighted that the above-mentioned measures, among others, have to be carried out within the context where social and health care policies offer support for early mothering experiences and their influence on parenting and infant feeding.⁵³

Conclusion

The association between breastfeeding and mother–infant relationships is complex and the manner in which breastfeeding is carried out would seem to be a decisive factor, influenced in turn by additional variables that should also be taken into account. Within the complex early relationships between a mother and a child, certain indicators associated to characteristics of the mother, such as maternal sensitivity, seem to be associated to breastfeeding, although no defined cause–effect associations can be established.

Certain additional maternal characteristics should be taken into account to understand the associations between breastfeeding and early relationships, such as self-efficacy about breastfeeding and autonomous breastfeeding motivation, as these have facilitating effects upon the relationship. The negative effect of identified breastfeeding regulation deserves special mention, because of its influence on the mother–child bond. This would help to understand the negative effect of family, health, and social pressures toward breastfeeding.

Regarding the variables associated to the child's behavior, breastfeeding is associated to secure attachment, especially when breastfeeding is continued up to 6 months after childbirth, although the underlying mechanisms are unknown. Breastfeeding favors optimal conditions for strengthening the mother–child bond, such as increased contact with the baby through feeding, and the need to read the baby's cues, especially during the first 3 months of life. Nevertheless, these conditions do not necessarily guarantee a mother–child bond, as this also depends on the quality of the bond being established through feeding.

As part of the complex associations between breastfeeding and early relationships, it is necessary to refer to the role that different variables play, including sociodemographic variables, early characteristics of the environment, maternal emotions, and the observed mother–child relationship quality. In summary, breastfeeding seems to be associated to mother–child relationships as long as it is under certain circumstances that favor the bond.

Authors' Biographies

C.P., psychologist, PhD, is a professor and a researcher at the Health Sciences Faculty (Rey Juan Carlos University). She is a specialist in the study of personality during pregnancy, from the perspective of Health Psychology.

P.C., psychologist, PhD Student, is a professor and a researcher at the Health Sciences Faculty (Rey Juan Carlos University). She is currently preparing her PhD thesis about the influence of psychosocial factors during pregnancy.

Footnotes

Disclosure Statement

No competing financial interests exist.

Funding Information

This work was funded by the Health Research Fund (Fondo de Investigaciones Sanitarias, FIS), grant number PI07/0571 from the Instituto de Salud Carlos III (Spain).

References

Netzer-Tomkins

, Rubin

, Ephros

. Breastfeeding is associated with decreased hospitalization for neonatal fever. Breastfeed Med, 2016; 11:218–221.

Binns

, Lee

, Low

. The long-term public health benefits of breastfeeding. Asia Pac J Public Health, 2016; 28:7–14.

Carter

CS.

Oxytocin pathways and the evolution of human behavior. Ann Rev Psychol, 2014; 65:17–39.

Feenstra

, Kirkeby

, Thygesen

, et al. Early breastfeeding problems: A mixed method study of mothers' experiences. Sex Reprod Healthcare, 2018; 16:167–174.

Hunt

, Thomson

. Pressure and judgement within a dichotomous landscape of infant feeding: A grounded theory study to explore why breastfeeding women do not access peer support provision. Matern Child Nutr, 2017; 13:e12279.

Jansen

, de Weerth

, Riksen-Walraven

. Breastfeeding and the mother–infant relationship: A review. Dev Rev, 2008; 28:503–521.

Moher

, Liberati

, Tetzlaff

, et al. Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement. Ann Intern Med, 2009; 151:264–269.

Bowlby

EJM.

Attachment and Loss, Vol. 1: Attachment. London: Hogarth Press, 1969.

Levine

, Zagoory-Sharon

, Feldman

, et al. Oxytocin during pregnancy and early postpartum: Individual patterns and maternal–fetal attachment. Peptides, 2007; 28:1162–1169.

10.

Ainsworth

, Blehar

, Waters

, et al.

Patterns of Attachment: A Psychological Study of the Strange Situation.

U.S.A.: Lawrence Erlbaum Associates, 1978.

11.

Reed

, Cook

, Beckman

, et al. Association between funding and quality of published medical education research. J Am Med Assoc, 2007; 298:1002–1009.

12.

Smith

, Learman

. A plea for MERSQI: The medical education research study quality instrument. Obstet Gynecol, 2017; 130:686–690.

13.

Friese

Qualitative Data Analysis with ATLAS.ti. London: SAGE Publications Ltd., 2012.

14.

Glaser

, Strauss

. The Discovery of Grounded Theory. Strategies for Qualitative Research. New Brunswick (USA): AldineTransaction, 1967.

15.

Whittemore

, Knafl

. The integrative review: Updated methodology. J Adv Nurs, 2005; 52:546.

16.

Abuhammad

SH.

The Possible Effect of Breastfeeding on Maternal Sensitivity During the First Year of Life [Doctoral Dissertation].

The University of Wisconsin-Milwaukee, Milwaukee, WI, 2016.

17.

Muller

ME.

A questionnaire to measure mother-to-infant attachment. J of Nurs Meas, 1994; 2:129–141.

18.

Cetisli

, Arkan

, Top

. Maternal attachment and breastfeeding behaviors according to type of delivery in the immediate postpartum period. Rev Assoc Méd Bras, 2018; 64:164–169.

19.

Yenal

, Okumuş

. Reliability of LATCH breastfeeding assessment tool. J Educ Res Nursing, 2003; 5:38–44.

20.

Gibbs

, Forste

, Lybbert

. Breastfeeding, parenting, and infant attachment behaviors. Matern Child Health J, 2018; 22:579–588.

21.

Andreassen

, Fletcher

Early Childhood Longitudinal Study, Birth Cohort (ECLS-B): Psychometric Report for the 2-Year Data Collection. Methodology Report. NCES 2007-084. National Center for Education Statistics, 2007.

22.

Nord

, Edwards

, Andreassen

, et al.

Early Childhood Longitudinal Study, Birth Cohort (ECLS-B), user's manual for the ECLS-B longitudinal 9-month–2-year data file and electronic codebook (NCES 2006–046).

National Center for Education Statistics, 2006.

23.

Keller

, Medved

, Armano

. The influence of maternal personality and risk factors for impaired mother-infant bonding on breastfeeding duration. Breastfeed Med, 2016; 11:532–537.

24.

Thompson

ER.

Development and validation of an international English big-five mini-markers. Pers Individ Dif, 2008; 45:542–548.

25.

Brockington

, Fraser

, Wilson

. The Postpartum Bonding Questionnaire: A validation. Arch Womens Ment Health, 2006; 9:233–242.

26.

Kestler-Peleg

, Shamir-Dardikman

, Hermoni

, et al. Breastfeeding motivation and self-determination theory. Soc Sci Med, 2015; 144:19–27.

27.

Ryan

, Deci

. Self-determination theory and the facilitation of intrinsic motivation, social development and well-being. Am Psychol, 2000; 55:68–78. https://dx-doi-org.web.bisu.edu.cn/10.1037/0003-066X.55.1.68.

28.

Watson

, Clark

, Tellegen

. Development and validation of brief measures of positive and negative affect: the PANAS scales. J Personality Soc Psychol, 1988; 54:1063–1070. https://dx-doi-org.web.bisu.edu.cn/10.1037/0022-3514.54.6.1063.

29.

Cox

, Holden

, Sagovsky

Detection of postnatal depression. Development of the 10-item Edinburgh postnatal depression Scale. Br J Psychiatry, 1987; 150(6):782–786. https://dx-doi-org.web.bisu.edu.cn/10.1192/bjp.150.6.782.

30.

Diener

, Emmons

, Larsen

, et al. The satisfaction with life scale. J Personality Assess 1985;49:71–75. https://dx-doi-org.web.bisu.edu.cn/10.1207/s15327752jpa4901_13.

31.

Johnston

, Mash

. A measure of parenting satisfaction and efficacy. J Clin Child Psychol, 1989; 18:167–175. https://dx-doi-org.web.bisu.edu.cn/10.1207/ s15374424jccp1802_8.

32.

Beinart-Babayoff

Change in Maternal Attachment Style: The Role of Self-efficacy, Perceived Social Support, Antepartum & Postpartum Mental Health Status [Doctoral dissertation]. Tel-Aviv University, Tel-Aviv, 2012.

33.

Brennan

, Clark

, Shaver

. Self-report measurement of adult attachment: an integrative overview. In: Simpson JA, Rholes WS. (Eds.), Attachment Theory and Close Relationships. Guilford, New York, 1998, pp. 46–76.

34.

Kielbratowska

, Kazmierczak

, Michalek

, et al. Temperament and the mother–infant dyad: Associations with breastfeeding and formula feeding with a bottle. Infant Ment Health J, 2015; 36:243–250.

35.

Matczak

, Jaworska

, Ciechanowicz

, et al. Children Development Scale DSR [Manual]. Warsaw: Psychological Test Laboratory of the Polish Psychological Association, 2007.

36.

Wolke

, James-Roberts

. Multi-method measurement of the early parent-infant system with easy and difficult newborns. Adv Psychol, 1987; 46:49–70.

37.

Brazelton

, Nugent

Neonatal Behavioral Assessment Scale.

Cambridge University Press, 1995.

38.

Kronborg

, Harder

, Hall

. First time mothers' experiences of breastfeeding their newborn. Sex Reprod Healthcare, 2015; 6:82–87.

39.

de Lauzon-Guillain

, Wijndaele

, Clark

, et al. Breastfeeding and infant temperament at age three months. PLoS One, 2012; 7:e29326.

40.

Gartstein

, Rothbart

. Studying infant temperament via the Revised Infant Behavior Qusestionnaire. Infant Behav Dev, 2003; 26:64–86.

41.

Liu

, Leung

, Yang

. Breastfeeding and active bonding protects against children's internalizing behavior problems. Nutrients, 2013; 6:76–89.

42.

Moffit

TE.

Juvenile delinquency and attention deficit disorder: Boys' developmental trajectories from age 3 to age 15. Child Dev, 1990; 61:893–910.

43.

Liu

, Cheng

, Leung

. The application of the preschool child behavior checklist and the caregiver-teacher report form to mainland Chinese children: Syndrome structure, gender differences, country effects, and inter-informant agreement. J Abnorm Child Psychol, 2011; 39:251–264.

44.

Loke

, Chan

LKS

. Maternal breastfeeding self-efficacy and the breastfeeding behaviors of newborns in the practice of exclusive breastfeeding. J Obstet Gynecol Neonatal Nurs, 2013; 42:672–684.

45.

Matthews

MK.

Mothers' satisfaction with their neonates' breastfeeding behaviors. J Obstet Gynecol Neonatal Nurs, 1991; 20:49–55.

46.

Dennis

CL.

The Breastfeeding Self-Efficacy Scale: Psychometric assessment of the short form. J Obstet Gynecol Neonatal Nurs, 2003; 32:734–744.

47.

Papp

LM.

Longitudinal associations between breastfeeding and observed mother–child interaction qualities in early childhood. Child Care Health Dev, 2014; 40:740–746.

48.

Caldwell

, Bradley

. Home Observation for Measurement of the Environment. Little Rock: University of Arkansas at Little Rock, 1984.

49.

Smith

, Ellwood

. Feeding patterns and emotional care in breastfed infants. Soc Indic Res, 2011; 101:227–231.

50.

Tharner

, Luijk

, Raat

, et al. Breastfeeding and its relation to maternal sensitivity and infant attachment. J Dev Behav Pediatr, 2012; 33:396–404.

51.

Brockington

, Oates

, George

, et al. A screening questionnaire for mother-infant bonding disorders. Arch Womens Ment Health, 2001; 3:133–140.

52.

Thomson

, Ebisch-Burton

, Flacking

. Shame if you do–shame if you don't: Women's experiences of infant feeding. Matern Child Nutr, 2015; 11:33–46.

53.

Stearns

CA.

The embodied practices of breastfeeding: Implications for research and policy. J Women Polit Policy, 2013; 34:359–370.