Longitudinal Associations between the Quality of Mother-Infant
Interactions and Brain Development across Infancy
Annie Bernier, Susan D. Calkins, and Martha Ann Bell
Annie Bernier, Department of Psychology, University of Montreal; Martha Ann Bell, Department of
Psychology, Virginia Tech; Susan D. Calkins, Departments of Human Development and Family
Studies and Psychology, University of North Carolina at Greensboro
Abstract
The aim of this study was to investigate if normative variations in parenting relate to brain
development among typically-developing children. A sample of 352 mother-infant dyads came to
the lab when infants were 5, 10, and 24 months of age (final N = 215). At each visit, child resting
EEG was recorded. Mother-infant interactions were videotaped at the 5-month visit. The results
indicated that higher-quality maternal behavior during mother-infant interactions predicted higher
frontal resting EEG power at 10 and 24 months, as well as increases in power between 5 and 10
months, and between 10 and 24 months. These findings provide rare support for the hypothesis
that normative variation in parenting quality may contribute to brain development among
typically-developing infants.
Inspired by Greenough, Black, and Wallace's (1987) influential propositions pertaining to
the experience-dependent nature of brain development, many scientists have been intrigued
by the possibility that early relational experiences could influence children's brain
development. These ideas have gained in popularity with the advent of modern brain
imaging techniques that have enabled scientists to appreciate the remarkable plasticity of the
developing brain, especially during the first years of life, which are characterized by over-
production of synapses followed by a period of gradual pruning (Huttenlocher, 2002).
During this period, experience is considered to determine to a large degree which synaptic
connections persist and are strengthened by frequent use, and which are selectively
eliminated as a result of inactivity (Singer, 1995). In this context, there is a substantial
window of time for environmental input to influence the developing brain (Kolb et al.,
2012). Such observations have led many to suggest that early caregiving relationships should
be centrally implicated in children's brain development (e.g., Belsky & de Haan, 2011;
Cicchetti, 2002; Gunnar, Fisher, & the Early Experience Stress and Prevention Science
Network, 2006; Nelson, 2000). Empirical evidence supporting these claims is, however, still
remarkably scarce. Tackling this gap, this longitudinal study investigated the prospective
links between the quality of maternal behavior during early mother-infant interactions and
infant brain development.
Correspondence concerning this article should be addressed to Annie Bernier, Department of Psychology, University of Montreal, P.O.
Box 6128, Downtown Station, Montreal QC H3C 3J7. annie.bernier@umontreal.ca.
HHS Public Access
Author manuscript
Child Dev
. Author manuscript; available in PMC 2017 July 01.
Published in final edited form as:
Child Dev
. 2016 July ; 87(4): 1159–1174. doi:10.1111/cdev.12518.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
As one of the earliest, most intense, and most enduring experiences of childhood, the parent-
child caregiving relationship appears to be a prime candidate to account for environmentally-
driven individual differences in children's brain development. Decades of empirical research
have provided overwhelming support for the classic notion that early parent-child
relationships exert an exceptional influence on child development. As documented by
longitudinal studies spanning infancy to early adulthood (Fraley, Roisman, & Haltigan,
2013; Grossmann, Grossmann, & Waters, 2005; Jaffee, Caspi, Moffitt, Belsky, & Silva,
2001), by meta-analytic reviews (Fearon, Bakermans-Kranenburg, Van IJzendoorn, Lapsley,
& Roisman, 2010; Groh, Roisman, Van Ijzendoorn, Bakermans-Kranenburg, & Fearon,
2012; Pallini, Baiocco, Schneider, Madigan, & Atkinson, 2014), and by experimental studies
(Guttentag et al., 2014; Kochanska, Kim, Boldt, & Nordling, 2013), the quality of caregiving
relationships forecasts child outcomes as diverse as social and emotional adjustment (see
Thompson, 2008), moral development (Dunn, Brown, & Maguire, 1995), cognitive
functioning (Tamis-LeMonda, Bornstein, & Baumwell, 2001), sleep/wake cycles
(Bordeleau, Bernier, & Carrier, 2012), and sympathetic and parasympathetic response (Luijk
et al., 2010). Such pervasive effects on socio-emotional, cognitive and biological functioning
are often believed to transit through children's neural circuitry (Belsky & de Haan, 2011;
Gunnar, 2003).
In this context, the dearth of studies directly examining the prospective links between early
caregiving experiences and children's brain development is striking. Nearly all knowledge on
this crucial topic is based on the study of grossly inadequate environments, characterized by
neglect, abuse (Curtis & Cicchetti, 2007; De Bellis, 2001, 2005; Teicher, Tomoda, &
Andersen, 2006), or extreme relational deprivation in orphanages (Chugani et al., 2001;
Marshall & Fox, 2004; Rutter & O'Connor, 2004). These studies show the profound impact
that severely unfavorable conditions can have on normal brain development, from both
anatomical and functional standpoints. Hence, brain growth appears to be very susceptible to
extreme adversity.
It is expected that the converse be true as well, such that favorable environmental
experiences should have a positive impact on healthy brain development (e.g., Nelson &
Bloom, 1997). Attempts at “enrichment” activities for children in deprived environments
provide some support for this view. Children in Mauritius receiving an enriched preschool
intervention and institutionalized children in Bucharest placed into foster care appear to
exhibit some changes in brain electrical activity indicative of brain maturation, relative to
care-as-usual groups of children (Marshall, Reeb, Fox, Nelson, & Zeanah, 2008; Raine et al.,
2001). It is also expected that typical positive social influences, especially those embedded
within early caregiving relationships, should foster optimal brain development (e.g., Schore,
2001). Surprisingly however, there is almost no empirical work to date to support such
claims in the context of non-extreme or pathological parenting. The time appears to be ripe
to investigate whether normative variations in parenting affect brain development among
typically-developing children.
Animal research suggests that this may be the case: enriched maternal nurturance promotes
increased dendritic branching, enhanced neurogenesis, and greater dendritic spine density in
a number of brain areas (for reviews see Belsky & de Haan, 2011; Markham & Greenough,
Bernier et al. Page 2
Child Dev
. Author manuscript; available in PMC 2017 July 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
2004; Meaney, 2001). Reviewing such studies on maternal care in animal populations,
Francis and Meaney (1999) concluded that variation within the normative range of human
maternal care, much less dramatic than abuse or neglect, could have a profound influence on
normal brain development. Over fifteen years later, the dearth of research properly
investigating these questions with human populations is remarkable. Studies to date have
almost exclusively used distal proxies rather than direct measures of actual parenting, and/or
cross-sectional designs with concurrent or retrospective measures.
Studies using distal indicators such as family socio-economic status (e.g., Jednoróg et al.,
2012; Noble, Houston, Kan, & Sowell, 2012), stressful life events in the family (Luby et al.,
2013), or maternal depression (Ashman, Dawson, & Panagiotides, 2008; Diego, Jones, &
Field, 2010; Lupien et al., 2011) converge to suggest that such indices of familial risk predict
non-optimal brain development in children, whether considering structure or function.
Importantly, the authors of these studies argue that such distal factors have the power to
impact children's brain development because they are likely to influence the quality of
parent-child interactions, which in turn is presumed to be the key factor influencing
children's brain development. The current study addresses this key factor directly among
typically-developing children.
Recently there have been several reports of associations between maternal behaviors during
infancy and later childhood performance on executive function tasks (e.g., Bernier, Carlson,
Deschênes, & Matte-Gagné, 2012; Cuevas et al., 2014; Kraybill & Bell, 2013). Given that
executive function is inextricably linked to the prefrontal cortex (Stuss, 2011), these findings
provide further indirect evidence for a potential link between early parenting and subsequent
brain development, perhaps especially in the frontal areas. In light also of evidence that early
experiences affect prefrontal circuitry (Kolb et al., 2012), this study focuses on frontal brain
regions.
Other studies have used adults' retrospective accounts of their childhood experiences. Some
of these studies found that perceived maternal care in childhood was related to structural
brain measures in adulthood (Buss et al., 2007; Kim et al., 2010), whereas others found no
links (Narita et al., 2012). Of note, however, is that these studies were cross-sectional, thus
leaving open the possibility that the observed brain markers were in fact responsible for
adults' recollections of their childhood. This is of special concern in light of the fact that
retrospective accounts of childhood experiences are notoriously inaccurate (Henry, Moffitt,
Caspi, Langley, & Silva, 1994; Roisman, Haltigan, Haydon, & Booth-LaForce, 2014).
Other, mostly cross-sectional studies have used electroencephalography (EEG) to assess
regional brain activity among infants, along with measures of parent-infant relationships.
These studies focused on resting frontal EEG asymmetry, a well-documented marker of
emotional and motivational tendencies (Davidson, 2000), with right frontal asymmetry
typically associated with stress. Dawson, Klinger, Panagiotides, Spieker, and Frey (1992)
found no direct link between frontal asymmetry and concurrent mother-infant attachment
security, as assessed with the Strange Situation Procedure (SSP) when infants were between
11 and 17 months of age. In contrast, Dawson and colleagues (2001) observed that infants
aged 13–15 months who were insecurely attached to their mothers were more likely to
Bernier et al. Page 3
Child Dev
. Author manuscript; available in PMC 2017 July 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
exhibit right frontal EEG asymmetry recorded during a resting baseline condition prior to
the SSP assessment. Jones, McFall, and Diego (2004) found that infants of depressed
mothers who had been breastfed until their third month of life were less likely than their
bottle-fed counterparts to show the right frontal EEG asymmetry usually associated with
maternal depression. Diego, Field, Jones, and Hernandez-Reif (2006) further reported that
infants of depressed mothers exhibiting a withdrawn interactive style at 6 months showed
greater right frontal asymmetry in EEG recordings at 3 to 6 months of age than those of
depressed mothers characterized by an intrusive style. Hane and Fox (2006) assessed
mother-infant home interactions when infants were aged 9 months, and found relations to
infants' concurrent resting frontal EEG asymmetry when considering extreme groups of
maternal behavior (+/− 1 standard deviation above or below the mean): infants exposed to
very low-quality maternal behavior were more likely to show right frontal asymmetry. In a
subsequent study with the same sample, Hane, Henderson, Reeb-Sutherland, and Fox (2010)
found no significant association between the quality of 9-month maternal behavior and
subsequent resting frontal asymmetry at 3 years of age, when considering the whole
distribution of maternal behavior scores (an analysis not reported in Hane & Fox, 2006);
however, they did find relations between 3-year frontal EEG asymmetry and the extremes of
maternal behavior, similar to that found when infants were aged 9 months. It is unknown,
however, whether the 3-year results held above the previously documented associations at 9
months.
Despite some inconsistency in results, the findings of these studies are noteworthy in that
they provide rare evidence that the quality of parent-child relationships, as measured directly
and objectively, may in some circumstances relate to an important aspect of infants' brain
functioning: infant right frontal asymmetry reflects a disposition toward anxiety and
generally negative emotionality (Davidson, 2000). They may be, however, less telling with
respect to brain development per se, given their cross-sectional designs or lack of statistical
control for prior EEG data, and in that frontal asymmetry is considered to be a trait-like
disposition, presumed to reflect stable individual differences in affective style. On the other
hand, EEG power has a long tradition as a measure of brain development (see Bell, 1998;
Bell & Fox, 1994, for reviews). The EEG signal reflects post-synaptic activity and EEG
power (level of mean square microvolts of the EEG signal resulting from Fourier analysis)
reflects the excitability of groups of neurons (Bell & Cuevas, 2012). With respect to brain
development during infancy, more neural activity translates into higher EEG power. Frontal
EEG power has clear functional implications, in that it consistently relates to infant cognitive
performance, for instance working memory and inhibitory control (e.g., Bell, 2001; Cuevas,
Bell, Marcovitch & Calkins, 2012). Importantly for our purposes, frontal resting EEG power
is also considered indicative of brain
development
, given that it increases almost linearly
with age across infancy (Bell & Fox, 1992; Cuevas & Bell, 2011). Accordingly, an age-
related increase in resting frontal EEG power may be a strong indicator of brain
development during infancy, and thus constitutes a particularly relevant outcome to consider
when assessing parenting influences on infants' brain development. To our knowledge,
however, research has never examined relations between parenting and infant EEG power,
whether in cross-sectional or longitudinal designs. Addressing this gap was the primary aim
of the current study.
Bernier et al. Page 4
Child Dev
. Author manuscript; available in PMC 2017 July 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
The current study
This longitudinal study used an observational measure of the quality of maternal behavior
during mother-infant interactions when infants were aged 5 months, along with concurrent
and subsequent EEG recordings at 10 and 24 months, to investigate whether the quality of
maternal parenting behavior relates to infant brain development, as indicated by frontal
resting EEG power as well as age-related increases in frontal resting power. It was expected
that infants whose mothers showed higher-quality behavior while interacting with them at 5
months would show higher frontal EEG power at 10 and 24 months, as well as more
pronounced increases in power between 5 and 10 months, and between 10 and 24 months.
Method
Participants
Participants were part of an ongoing longitudinal study examining individual differences in
cognitive and emotion processes across early development. Data were collected from 352
mother-infant dyads (185 girls) recruited by two research locations (Blacksburg, VA;
Greensboro, NC), with each location recruiting half of the total sample. Dyads first came to
the research labs when infants were 5 months of age (
M
= 162 days,
SD
= 8 days). All
infants were born after at least 36 weeks of pregnancy, had a normal birth weight (> 5.5
pounds), and none suffered from a known developmental disorder diagnosed in early
childhood. Of the 352 infants, 272 were Caucasian, 48 African American, 15 multiracial, 2
Asian, 13 were other, and two did not report; with respect to ethnicity, 22 were Hispanic and
one did not report. Mothers and fathers were 29 and 32 years old on average (
SD
: 6 and 7
years) when infants were born. Sixty-five percent of mothers and 56% of fathers had a
college education or higher. Infants were recruited via commercial mailing lists, newspaper
birth announcements, and word of mouth. This sample was drawn from a larger pool of 410
dyads who participated in the lab visit at 5 months. Given the goals of the current study, we
excluded participants who had unusable EEG data or unusable/unavailable mother-infant
interactions.
Of the 352 dyads, 304 (86.4%; 155 girls) returned for a second laboratory visit when infants
were aged 10 months (
M
= 314 days,
SD
= 11 days). Finally, 215 of the original dyads
(61.2%; 116 girls) returned for a third lab visit when children reached 24 months of age (
M
= 25 months,
SD
= 20 days). Eighteen of these 215 dyads had not taken part in the 10-month
visit; therefore, complete data at all three time points are available for 197 dyads (105 girls).
Attrition was mainly due to families moving out of the local regions. Mothers were paid $50
for each visit and children received a small toy.
Attrition analyses revealed that the 197 families with complete data differed from the 155
families for whom data were missing at the 10-month visit, 24-month visit, or both, on
mother's age at infant birth. Mothers who completed all visits were older (
M
= 30.16) at
their infant's birth than mothers who failed to complete one or both follow-up visits (
M
=
28.45;
t
(350) = 2.93,
p
= .004). These groups did not differ on child sex (
p
= .73), maternal
education (
p
= .13), or any of the four dimensions of maternal parenting behaviors assessed
at 5 months (described below;
p
's between .68 and .77).
Bernier et al. Page 5
Child Dev
. Author manuscript; available in PMC 2017 July 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript