Mother-Infant Interaction and Child Brain Morphology: A Multidimensional Approach to Maternal Caregiving Behavior

• In Event: 1-181 - Normative Variation in Early Caregiving: Longitudinal Links to Brain Structure and Function

Integrative Statement

Studies of institutionalized and maltreated children suggest that particular regions of the brain may be especially sensitive to caregiving influences, notably the hippocampus and amygdala (Tottenham & Sheridan, 2010). Both the hippocampus and amygdala are centrally implicated in socio-emotional processing early in life. Consequently, they can be expected to be activated during, and thus influenced by, daily interactions between infants and their caregivers, with the associated repeated patterns of activation having long-term consequences for children’s neural development. This study investigated whether the quality of maternal caregiving behavior in infancy predicted the volumes of the amygdala and the hippocampus in late childhood.

Given increasing evidence that maternal behavior is a multidimensional construct, three empirically derived dimensions were considered (Bailey et al., 2016): Cooperation (mothers’ ability to accurately interpret infant cues and adjust the interaction correspondingly), Positivity (maternal positive affect while interacting with the infant), and Accessibility (consistent attentiveness toward the infant). These three dimensions were assessed among 33 mothers of one-year-old children with the Maternal Behavior Q-Sort, based on observations conducted throughout a 1.5-hour home visit. When children reached 10 years of age, they underwent structural magnetic resonance imaging using a 3D T1-weighted protocol on a Siemens 3T scanner. Volumes of the hippocampus and amygdala were obtained using CAT12 (in SPM12) with a label atlas based on manually segmented whole-brain parcellation (Neuromorphometrics).

The three dimensions of maternal behavior were entered simultaneously in regression equations as predictors of each region of interest: left amygdala, right amygdala, left hippocampus, right hippocampus. Child sex, age, and total intracranial volume, as well as maternal education and family income were covaried.

As displayed in Table 1, maternal positivity during mother-infant interactions uniquely and negatively predicted both left (Beta = -.54, p = .003) and right hippocampal volumes (Beta = -.58, p < .001). Maternal accessibility uniquely predicted a smaller right amygdala volume (Beta = -.64, p < .001). Volume of the left amygdala was not predicted by maternal behavior, and maternal cooperation showed no significant links to hippocampus and amygdala volumes.

Accessibility focuses on mothers’ awareness of the infant even when otherwise occupied, which likely allows mothers to intervene rapidly when needed. If an infant’s mother reliably intervenes when perceived threats arise in the environment, the infant amygdala’s threat-detection function is less intensely recruited, which may protect the amygdala from premature structural development (Tottenham, 2012). With respect to the hippocampus, it is involved in the construction of representations of relationships (relational memory theory; Cohen & Eichenbaum, 1993). This is reminiscent of the notion of “internal working models” (IWM) of attachment, which are mental representations of the quality of parent-child interactions. Thus, the link observed here between maternal positivity, the behavioral dimension with the most overt emotional content of the three assessed here, and hippocampal volume raises the possibility that the hippocampus is implicated in the neural network subsuming IWM of the emotional quality of mother-infant relationships.

Overall, the results suggest that a multidimensional approach to maternal behavior is a fruitful way to advance research in this area.

Authors

• Annie Bernier, University of Montreal
  Presenting Author

• Élizabel Leblanc, University of Montreal
  Non-Presenting Author

• Veronique Daneault, Center for Advanced Research in Sleep Medicine, Montreal Sacré-Cœur Hospital, Quebec, Canada
  Non-Presenting Author

• Heidi Bailey, University of Guelph
  Non-Presenting Author

• Miriam H. Beauchamp, University of Montreal
  Non-Presenting Author