Poster #150 - The Beneficial Effect of Breastfeeding on Bonding and Attachment: a Meta - Analysis

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Parenting & Parent-Child Relationships

Integrative Statement

The World Health Organization recommends 6 months of exclusive breastfeeding for the infant to benefit from positive effects of breastfeeding on child physical health (2003) such as reduced risk of infectious diseases (Duijts, Jaddoe, Hofman & Moll, 2010) and type 2 diabetes (Owen, Martin, Wincup, Smith & Cook, 2006). Furthermore, for mothers, breastfeeding has a protective effect against breast cancer (Freund, Mirabel, Annane & Mathelin, 2005). Several meta-analyses demonstrated the positive association between breastfeeding and later maternal or infant health (Oddy, 2001; Harder, Bergman, Kallischnigg & Plagemann, 2005), but only one article reviewed the association between breastfeeding and the mother-infant relationship (Jansen, de Weerth & Riksen-Walraven, 2008). This review concluded that the scientific evidence for a positive relationship between breastfeeding and the mother-infant relationship was insufficient. Nevertheless, a review by Schulze and Carlisle (2010) found that numerous articles list bonding and attachment as one of the benefits of breastfeeding. Our goal was to create a recent summary on the association between breastfeeding and the mother-infant relationship, which was conceptualized as the maternal bond towards her infant and infant attachment towards its mother.
A literature search was carried out in Embase, Medline Ovid, Web of Science, Cochrane CENTRAL, CINAHL EBSCOhost, PsycINFO Ovid, ERIC EBSCOhost and Google Scholar for papers combining the terms breastfeeding with attachment or bonding. Studies were screened on eligibility by two independent researchers. The initial search yielded 5768 articles after removal of duplicate records. After screening titles and abstracts, 17 studies were selected for full-text, of which 11 were relevant for our synthetization.
A meta-analysis showed a combined effect size of r = 0.13 , 95% CI [0.04 – 0.22] and a 95% PI [-0.26 – 0.48]. The combined effect size only reflects a very low positive correlation between breastfeeding and mother-infant interaction. Moreover, the I2 statistic provides a percentage of 95, suggesting that studies in this meta-analysis are too heterogenetic to consider an overall effect. After subgroup analysis of the studies on attachment and bonding separately, the studies remained too heterogenic for interpretation. As can be seen in Table 1, only one study confirms the positive association between breastfeeding and attachment (whereas two do not) and four confirm the positive association with bonding (whereas three do not). Differences between studies may be explained by the vast variety of measurement instruments used in these studies. In fact, all eleven studies used a different instrument to measure attachment and bonding. Furthermore, it is possible that the association between breastfeeding and the mother-infant relationship is confounded by other factors, such as social economic status, stress and maternal mental health.
Based on our synthesis, statements about a positive association between breastfeeding and maternal bonding and infant attachment are not unambiguously supported by empirical evidence. More research is necessary to warrant recommendations for breastfeeding on grounds of beneficial effects on the mother-infant relationship. Therefore, we will expand our search to include studies on maternal sensitivity, as well as increasing our efforts to find additional studies on bonding and attachment through snowballing and expert consultation.

Authors

• Marije van der Hulst, Erasmus MC
  Presenting Author

• Rianne Kok, Department of Psychology, Education, and Child Studies, Erasmus University Rotterdam
  Non-Presenting Author

• Willemijn Maria van Eldik, Erasmus University Rotterdam
  Non-Presenting Author

• Peter Prinzie, Erasmus University Rotterdam
  Non-Presenting Author