You are here

Breastfeeding and Its Relation to Maternal Sensitivity and Infant Attachment

Breastfeeding for longer was associated with more maternal sensitivity, more attachment security, and less attachment disorganization, but bottle-feeding did not necessarily harm the mother-infant relationship.

[popup title="Sample, Objective, Variables Measured, Study Design" format= "Default hover" text='


  • N: 675
  • Subject Ages: Children 2 to 6 months old at T1, mean maternal age at intake was 32 years
  • Location: Netherlands, Rotterdam
  • SES: Not available
  • Eligibility: Children born between February 2003 and August 2005 with two parents and four grandparents born in the Netherlands
  • Additional:
    • White Dutch
    • 49.2% girls
    • 63.8% of the children were born with spontaneous deliveries
    • Average birth weight was 3517 g, 3.7% of children had a low birth weight (2,500 g or less)
    • Mean five-minute Apgar score was 9.6, and only 0.9% had a score at or below 7, which is considered pathological
    • Educational level was generally high - 35.7% of the mothers had completed education beyond the bachelor’s degree


  1. Investigation into whether breastfeeding duration during the first 6 months is associated with maternal sensitive responsiveness, attachment security and attachment disorganization in a large prospective birth cohort.
  2. Examined the potential mediating or moderating roles of maternal sensitive responsiveness and maternal oxytocin receptor genotype.

Variables Measured, Instruments Used

  • Breastfeeding -
    • initiation data from delivery reports
    • continuation data from postal questionnaires at 2 and 6 months after birth
  • Maternal sensitive responsiveness - observed using Ainsworth’s rating scales
  • Attachment security and disorganization - the Strange Situation Procedure
  • OXTR Genotype - mothers were genotyped for polymorphisms in the oxytocin receptor gene, OXTR

Design—Cohort, prospective



  1. Longer duration of breastfeeding was associated with more maternal sensitive responsiveness, more attachment security and less attachment disorganization.
  2. Longer duration of breastfeeding was not related to the risk of insecure-avoidant or insecure-resistant versus secure attachment classification but was predictive of a lower risk of disorganized versus secure attachment classification.
  3. Maternal sensitive responsiveness did not mediate the associations.
  4. Maternal oxytocin receptor genotype was not a significant moderator.
  5. Breastfeeding initiation: About 10% of the mothers never initiated breastfeeding, about 22% started breastfeeding but stopped before the child was 2 months old, 36% continued beyond the second month but stopped before the child was 6 months old and 31% of the participating mothers breastfed their child for at least 6 months.
  6. When the infant was 2 months old, 35% of the mothers were bottle-feeding the child, 18% were combining bottle-feeding and breastfeeding, and 47% exclusively breastfed.
  7. Attachment classification distributions: 49.8% secure, 13.2% insecure-avoidant, 14.7% insecure-resistant and 22.4% disorganized.


  • Inability to make inferences about the causality of the associations of breastfeeding duration with maternal sensitive responsiveness, attachment security and attachment disorganization found in this study. For example, it may be that more sensitive women are more likely to breastfeed their child (and to continue breastfeeding to an older age) than less sensitive women. It has also been suggested that maternal personality characteristics account for the association between breastfeeding duration and maternal sensitivity, affecting both the decision to breastfeed (and continue to breastfeed) and maternal behavior toward the child. The continuation of breastfeeding may also be influenced by external factors, such as policy regarding maternity leave.
  • This study population was relatively homogeneous: Only white Dutch women and children were eligible for this study, socioeconomic status (SES) was relatively high and the level of other risk factors, such as parental psychopathology, was low. This reduces the effect of potential confounders but may limit the generalizability of our results to Western low-risk populations.
  • Even in dyads who never breastfed, maternal sensitive responsiveness was high and attachment disorganization scores were well below the clinical cutoff of 5. Categorical analysis revealed no differences in attachment classification. These results concur with earlier findings that although breastfeeding predicted better interactions between mothers and their 1-year-olds, bottle-feeding did not harm the infant-mother relationship.
  • It cannot be concluded that there was a mediating role of maternal sensitivity between breastfeeding duration and attachment security from the findings, because maternal sensitivity was not related to attachment security in this study. This lack of an association between maternal sensitive responsiveness and attachment security was unexpected but not inconsistent with earlier findings. The somewhat unusual setting of the assessments used in this study (infant sitting on the mother’s lap) might partly explain the lack of an association.