Brain and Cognition 124 (2018) 47–56

Contents lists available at ScienceDirect

Brain and Cognition

journal homepage: www.elsevier.com/locate/b&c

The neural correlates of attachment security in typically developing children T

a,b b,c,d e f a,⁎
Eun Jung Choi , Margot J. Taylor , Soon-Beom Hong , Changdai Kim , Soon-Hyung Yi
a
Department of Child and Family Studies, College of Human Ecology, Seoul National University, Seoul, South Korea
b
Department of Diagnostic Imaging, Hospital for Sick Children, Toronto, Canada
c
Neuroscience & Mental Health Program, Hospital for Sick Children Research Institute, Toronto, Canada
d
Department of Psychology and Medical Imaging, University of Toronto, Toronto, Canada
e
Division of Child and Adolescent Psychiatry, Department of Psychiatry, Seoul National University Hospital, Seoul, South Korea
f
Department of Education, College of Education, Seoul National University, Seoul, South Korea

A R T I C LE I N FO A B S T R A C T

Keywords: This study investigated neural correlates of children’s attachment security using functional magnetic resonance
Attachment security imaging. Fifty-one boys’ attachment styles (age mean = 9.5 years, SD = 0.61) were assessed with the Separation
fMRI Anxiety Test (SAT). We created an fMRI version of the SAT to activate children’s attachment system in fMRI
Striatum environment and contrasted two conditions in which children were instructed to infer the specific feeling of the
Dorsolateral prefrontal cortex
boy in the picture or to identify objects or physical activities. In the final fMRI analysis (N = 21), attachment
Amygdala
security could be detected at the neural level corresponding to the behavioural differences in the attachment
interview. Securely attached children showed greater activation in the frontal, limbic and basal ganglia area
which included the dorsolateral prefrontal cortex, amygdala, cingulate cortex and striatum, compared to other
children who had lower quality of attachment. These regions have a key role in socio-emotional information
processing and also represent a brain network related to approach and avoidance motivation in humans.
Especially the striatum, strongly linked to reward processing underpinning social approach and avoidance
motivation, showed the largest effects in these differences and also positively correlated with emotional open-
ness scores in SAT. This suggests that the quality of attachment configures the approach and avoidance moti-
vational system in our brain mediated by the striatum.

1. Introduction that children rely on the caregiver’s availability as a source of comfort

1.1. Attachment security
Attachment is a deep and strong emotional bond that connects an
infant to a caregiver at an early stage of life and it is characterized by
specific approach behaviours in children to seek proximity with the
attachment figure, especially when upset or threatened (Ainsworth,
1973; Bowlby, 1969). It is almost universal that infants develop at-
tachment relationships, but there are individual differences in how ef-
fectively the infants and children can use their caregivers as a source of
comfort in the face of danger or threats from the environment and how
balanced they are between exploration and seeking proximity to the
caregiver in various situations. Attachment security, that refers to these
individual differences in the quality of the attachment relationship, is
well-established at a behavioural level and has been classically divided
into two categories: “secure” and “insecure” attachment relationships
(Ainsworth et al., 2015; Bowlby, 1973). Secure attachment indicates
and protection when they feel unsettled or fearful of something in the
environment, while those with insecure attachment have not experi-
enced consistent availability from their caregivers and become anxious,
expressing fear or anger that their caregivers are not responsive when
needed (Ainsworth et al., 2015). Insecure attachment style can take the
form of anxious attachment and avoidant attachment (Ainsworth et al.,
2015). Attachment anxiety is predicted by the receipt of unreliable or
unpredictable caregiving, whereas experiences of rejection by care-
givers predict the development of an avoidant attachment style. In-
dividuals high in avoidant attachment style dismiss the importance of
attachment bonds, while anxiously attached individuals are hypervi-
gilant for signs of social rejection, and readily admit their longing for
improved attachment relationships. Over time, each group tends to
develop a diverse and mixed set of profiles, but a number of long-
itudinal studies on attachment have shown that these differences in
attachment security are stable across the life span from infancy to
adulthood (Waters, Merrick, Treboux, Crowell, & Albersheim, 2000)

⁎
Corresponding author at: Dept. of Child Development & Family Studies, Seoul National University, Seoul, South Korea and Dept. of Diagnostic Imaging, Hospital for Sick Children,
Toronto, Canada.
E-mail addresses: eunjchoi3@gmail.com (E.J. Choi), ysh@snu.ac.kr (S.-H. Yi).

https://doi.org/10.1016/j.bandc.2018.04.003
Received 11 December 2017; Received in revised form 14 April 2018; Accepted 15 April 2018
0278-2626/ © 2018 Published by Elsevier Inc.
E.J. Choi et al.
and even extend across generations (Benoit & Parker, 1994).
One reason that attachment security has a prolonged effect on an
individual’s life is that attachment is basically supported by a physio-
logical system that is biologically predisposed and selected during
evolution for survival (Bowlby, 1969, 1982). The attachment beha-
vioural system can be viewed as a physiological adaptation and
homeostasis, such that attachment security can be understood as a re-
sult of adapting the attachment behavioural system to the caregiving
environment. A second reason that attachment security is stable over
time is that from interactions with primary caregivers, children develop
mental representations of the self and others, called “internal working
models”, which are internalized in one’s mind and utilized as a source
of how to react to socio-emotional cues in close relationships and more
generalized social contexts (Bretherton & Munholland, 2008). How-
ever, the neural substrates underlying this stable attachment security
remain largely unknown.
1.2. The neural basis of attachment security
There is growing empirical evidence that individual differences in
attachment security are associated with different neurobiological
functions in the processing of perceptual information (Cohen & Shaver,
2004; Mikulincer & Shaver, 2007), especially in emotional information.
Regions in prefrontal, limbic and basal ganglia are thought to be re-
sponsible for individual differences in attachment security. However,
the amygdalae also are often related to attachment security. During
processing various emotional information such as social feedback in
gaming situations (Vrticka, Lordier, Bediou, & Sander, 2014), emo-
tional faces (Norman, Lawrence, Iles, Benattayallah, & Karl, 2014;
Redlich et al., 2015; Vrticka, Andersson, Grandjean, Sander, &
Vuilleumier, 2008), attachment-related stress (Lemche et al., 2006; Liu,
Ding, Lu, & Chen, 2017), emotional adjectives (Debbané et al., 2017),
emotional regulation (Moutsiana et al., 2014), or mother’s face
(Tottenham, Shapiro, Telzer, & Humphreys, 2012), amygdalae showed
different patterns of activity in individuals with different attachment
styles. These results suggested that the different sensitivity for emo-
tional cues may be modulated by attachment security in the amygdalae.
The basal ganglia serve a wide range of functions, including motor,
cognitive, motivational, and emotional processes (Arsalidou, Duerden,
& Taylor, 2013). Consistent with this broad-reaching involvement in
brain function, basal ganglia dysfunction has been implicated in nu-
merous neurological and psychiatric disorders. Recent research re-
vealed a critical role for brain reward systems, involving the basal
ganglia, in the biological underpinnings of attachment security, as
neural responses to facial expression (Donges et al., 2012; Vrticka et al.,
2008), reward processing (Takiguchi et al., 2015), or the perception of
mother’s face (Minagawa-Kawai et al., 2008). There is mounting evi-
dence for a role of reward circuits and reinforcing processes in social
approach and bonding in maternal and romantic love (Aron et al.,
2005; Fisher, Aron, & Brown, 2005; Lenzi et al., 2008; Nitschke et al.,
2004; Ranote et al., 2004; Sander, Frome, & Scheich, 2007).
Most of research on the neural basis of attachment security, how-
ever, has studied adult populations, although attachment security is
formed at an early stage of life. Investigating the neural underpinnings
of attachment security in childhood would allow us to determine if the
adult model applies to younger cohorts, as well as determining possible
developmental differences in this complex social behaviour.
Furthermore, most research has used relatively generalized emotional
stimuli such as facial expressions to examine functional regions related
to attachment security, but attachment security is classically assessed
by activating one’s attachment system through presenting negative at-
tachment-related stimuli, such as strange situations in infancy and se-
paration pictures in childhood and adolescence. This leaves open the
question of whether or not the differences in brain activity during
processing facial expressions would reflect the neural basis of attach-
ment security, or would be simply distinct neural responses to facial
48
Brain and Cognition 124 (2018) 47–56
expression in individuals who have a different styles of attachment.
One research group studied the attachment system specifically in
the fMRI environment, to examine the functional neuroanatomy of at-
tachment security. Buchheim et al. (2006) used attachment-related
scenes to activate the attachment system in adult patients with bor-
derline personality disorder (BPD). The BPD patients showed sig-
nificantly more anterior midcingulate cortex activity in response to
pictures which depicted characters facing attachment threats alone and
their unresolved attachment was associated with increasing amygdala
activation (Buchheim et al., 2006). Here we employed this paradigm
with healthy children to determine the functional neuroanatomy of
attachment security and to investigate differences in brain activity in
children with different levels of attachment security.
2. Material and methods
2.1. Participants
Fifty-one boys were recruited from elementary schools (age
mean = 9.5 years, SD = 0.61). After receiving informed consent,
trained examiners visited their home and assessed attachment security
using the Separation Anxiety Test (SAT; Hansburg, 1972; Resnick,
1993). The SAT is a semi-projective interview using separation pictures
to assess attachment from preschool-aged children to adolescents. All
interviews were audiotaped, transcribed verbatim, and coded according
to the SAT manual (Resnick, 1993). The SAT rating scales include nine
subscales (e.g., Emotional openness, Devaluing of attachment, Self-
blame, Resistance/Withholding, Preoccupied anger, Displacement of
feelings, Anxiety, Coherence of transcript, Solutions), and assign each
child to one of the attachment classifications based on the profile in
each subscale (e.g., Secure, Insecure_avoidant, Insecure_preoccupied).
As a result, twenty-nine boys (56.9%) were classified as secure at-
tachment and twenty-two boys (43.1%) were classified as insecure at-
tachment. Secure attachment is prevalent in diverse populations and
children in this group show a diverse set of profiles that are considered
more qualitative rather than quantitative (Diamond & Marrone, 2003).
Based on the SAT coding system, only fourteen boys in secure group
could be assigned into the prototypical “securely attached” category
with very high scores for emotional openness compared to the others in
secure group who revealed restricted feelings. Insecure attachment is
represented by two distinct patterns, avoidant or ambivalent type. In
our study, nineteen out of twenty-two in the insecure group showed
avoidant attachment style. All subjects who declined the brain scan or
had a history of psychiatric or neurological illness, left handedness, or
metal in their body were excluded. Additionally, eight subjects were
also excluded from more than 4 mm head motion during the scan. Fi-
nally, 21 boys were included: seven boys were in the prototypical se-
curely attached group, another seven were securely attached but re-
stricted in expressing feelings and seven boys were in the insecurely
attached group, six of whom had avoidant attachment type. Demo-
graphic information is presented in Table 1. Parents’ educational level
and social economic status were similar across all participants. All
parents reported that the main caregiver of the child during the first
three years was the mother. This study was approved by the institu-
tional review board for human subjects at the Seoul National Uni-
versity. All children and their parents provided written informed con-
sent prior to study entry.
2.2. Stimulus materials and procedure
We constructed an fMRI version of the SAT to activate children’s
attachment system and to elicit mental engagement with attachment-
related experiences in the MRI environment. The stimulus material
consisted of attachment-related pictures which originally came from
other versions of SAT, the Adult Attachment Projective Picture System
(AAP), or neuroimaging studies that measured neural response of the
E.J. Choi et al. Brain and Cognition 124 (2018) 47–56

Table 1
Demographics and neuropsychological assessments.
Attachment security type Test statistics

Prototypical secure (Mean ± SD) Secure but restricted (Mean ± SD) Insecure (Mean ± SD)

Children’s age 9.60 ± 0.52 9.71 ± 0.41 9.42 ± 0.78 F = 0.45

Father’s education level

University or more 7 (100%) 5 (71.4%) 4 (57.1%) χ2(2)=3.68
High school graduation 0 (0%) 2 (28.6%) 3 (42.9%)

Mother’s education level

University or more 7 (100%) 4 (5731%) 5 (71.4%) χ2(2)=3.68
High school graduation 0 (0%) 3 (42.9%) 2 (28.6%)
SES
High 1 (14.3%) 0 (0%) 1 (14.3%) χ2(2)=2.12
Middle 5 (71.4%) 6 (85.7%) 6 (85.7%)
Low 1 (14.3%) 1 (14.3%) 0 (0%)

Attachment Security Score 52.43 ± 2.70 51.29 ± 6.63 47.29 ± 7.57 F = 1.41

IQ (K-WISC-IV)
VCI 123.43 ± 12.38 116.71 ± 18.48 115.00 ± 12.95 F = 0.59
PRI 122.71 ± 12.72 112.86 ± 16.43 106.00 ± 20.42 F = 1.68
WMI 116.14 ± 12.93 117.57 ± 15.13 95.00 ± 12.15 F = 5.50*
PSI 94.43 ± 2.88 107.86 ± 12.84 96.50 ± 13.53 F = 3.14
FSIQ 121.43 ± 10.77 118.71 ± 15.06 105.83 ± 14.73 F = 2.37

CBCL total score 22.86 ± 19.44 22.00 ± 12.88 15.14 ± 10.06 F = 0.42

*
p < .05.

Table 2
Examples of the experimental conditions.
Attachment condition Control condition

Target stimuli

Neutral stimuli

activation of attachment system for adults (Buchheim et al., 2006,
2008). These pictures depicted attachment-related distress such as se-
paration, loneliness, death, or threat using black and white line drawing
and contained the essential components to recognize the context
without facial expressions. Each picture was combined with a question
“How does the child feel now?” to establish an attachment condition,
designed for subjects to focus on the emotional component. For the
control condition, the same pictures were combined with a different
question, asking about physical characteristics (e.g., “Where is the child
sitting now?”, “What is that beside the child?”, etc.). Thus, we could
compare directly the neural activations when children’s attachment
system was activated or not. Neutral stimuli were included as filler
items and consisted of pictures which portrayed adult-child dyads but
without an emotional component and were also combined with two
different questions in each condition (Table 2). To validate the stimuli,
a separate group of twenty-nine after-school-program students con-
ducted valence ratings between attachment-related scenes and neutral
scenes. The result of these ranking showed a significant difference be-
tween two sets of stimuli (t = −17.78, p < .001).
In the attachment condition, children were instructed to infer the
specific feeling of the boy in the picture while the scene was presented,
and then answer using a button box when the response screen came up.
In the control condition, children were instructed to identify objects or
activities according to the questions while the scene was presented, and
then to answer using button box when the response screen came up as
well. They had a training session prior to entering the scanner, so the
task and instructions were clearly understood. Each block was assigned
either attachment condition or control condition and it started with an
instruction screen for 5.4 s and then one of two targets and one neutral
stimuli were presented in random order for 6 s each followed by a re-
sponse screen for 3.8 s and a fixation cross for 1 s (see Fig. 1). Each run
had six blocks and it took about 5 min. We repeated this paradigm 3
times so the entire session lasted 15.4 min.

49
E.J. Choi et al.
Fig. 1. Experimental design.
2.3. Neuropsychological assessment
Neuropsychological measures were used to assess children’s IQ,
behavioural problems, and self-reported attachment security. A mea-
sure of IQ was obtained using the Korean version of Wechsler
Intelligence Scale for Children Fourth Edition (K-WISC-IV: Gwak, Oh, &
Kim, 2011), which included Verbal Comprehension Index (VCI), Per-
ceptual Reasoning Index (PRI), Working Memory Index (WMI), Pro-
cessing Speed Index (PSI). The Child Behavior Checklist (CBCL) was
also completed by a caregiver to identify behavioural problems in
children (Achenbach & Rescorla, 2001). We also used the Attachment
Security Scale (Kerns, Aspelmeier, Gentzler, & Grabill, 2001), a self-
reported questionnaire from children, to assess children’s perceptions of
attachment quality.
2.4. fMRI data acquisition
Images were acquired using Siemens Tim Trio 3.0 Tesla with 32
channel head coil. fMRI data were obtained using T2-weighted gra-
dient-echo EPI pulse sequences. Sponge padding was placed around the
head to minimize head movement and the lower jaw was fixed with
tape. Participants watched the monitor through a mirror on the head
coil, and held the button box in their right hand. MRI data were ac-
quired using the following parameters (T2-weighted gradient-echo EPI;
TR = 2700 ms; TE = 30 ms; flip angle = 90°, matrix = 64 × 64). Each
volume was constructed by 3 mm thick interleaved slices parallel to
anterior-posterior commissure line. High resolution T1-weighted
(MPRAGE) whole brain images were also obtained as an anatomical
underlay (T1-weighted images; voxel size = 1 × 1 × 1 mm;
TR = 1900 ms; TE = 2.36 ms; flip angel = 9°; matrix = 256 × 256).
2.5. fMRI data analysis
Using Statistical Parametric Mapping (SPM8: Wellcome Department
of Cognitive Neurology, Institute of Neurology, London, UK; www.fil.
50
Brain and Cognition 124 (2018) 47–56
ion,ucl.ac.uk/spm), all functional images for each participant were: (a)
realigned to correct for head motion; (b) spatially normalized into an
EPI image in SPM template for between-group analysis; and (c)
smoothed using a 8-mm full-width, half-maximum isotropic Gaussian
kernel. Following image conversion and preprocessing, the imaging
data were analyzed using SPM8. For each participant, condition effects
(attachment targets vs. control targets) were estimated according to the
general linear model, using a canonical hemodynamic response func-
tion. The resulting contrast images were entered into second-level
analyses using a random effects model to allow for inferences to be
made at the population level (Friston, Holmes, Price, Büchel, &
Worsley, 1999). We investigated the overall brain activation in each
group by one-sample t-test; (1) prototypical secure (PS) (2) secure but
restricted (RS) (3) insecure (IS). Then we used SPM full factorial model
with group (3) by condition (2). Comparisons between conditions in
whole-brain activation maps were first thresholded at uncorrected
p < .001 to see overall tendencies and we then applied a family wise
error correction p < .05. To examine any correlations between brain
activity and attachment rating scores, the ROIs were created using peak
MNI coordinates based on the results which showed group differences.
A 5-mm-diameter sphere was drawn surrounding the peak MNI co-
ordinates.
3. Results
3.1. Group characteristics
Table 1 contains the neuropsychological information for the three
groups of children. Children who were classified in secure attachment
tended to have higher score in the self-reported attachment security
ratings, but there was no significant group difference. Children who
were classified in insecure attachment showed significantly lower
working memory index (F = 5.50, p = 0.014) (see Table 1), but the full
scale IQ was not significantly different from other groups. There were
no significant group differences in behavioural problems.
E.J. Choi et al. Brain and Cognition 124 (2018) 47–56

Table 3 Table 4
Overall brain activation to the contrast conditions in prototypical secure group. Overall brain activation to the contrast conditions in insecure group.
Regions BA L/R Peak voxel (x,y,z) k z Regions BA L/R Peak voxel (x,y,z) k z

Attachment > Control Control > Attachment

Inferior frontal gyrus 45 L −50 20 2 27 4.37 Precuneus 19 R 22 −80 38 67 4.55
Inferior frontal gyrus 47 L −42 24 −6 4.02 Precuneus 7 R 22 −68 38 4.28
Caudate R 10 12 −2 33 4.24 Middle temporal gyrus 39 R 30 −68 26 4.26
Caudate R 10 4 2 3.72 Precuneus 31 L −26 −76 14 116 4.41
Putamen R 18 4 6 3.48 Middle temporal gyrus 19 L −34 −80 22 4.31
Anterior cingulate gyrus 24 L −6 36 14 117 4.05 Cuneus 7 L −22 −80 34 4.04
Dorsolateral prefrontal gyrus 8 L −6 48 46 4.03 Inferior temporal gyrus 37 L −54 −56 −14 8 3.59
Dorsolateral prefrontal gyrus 8 R 6 40 46 4.00
Amygdala 34 L −18 −8 −22 5 4.00 p < .001(uncorrected), k5, BA: Brodmann area. L/R:left/right, peak
Anterior cingulate gyrus 24 R 6 32 6 8 3.52 voxel = MNI xyz co-ordinates, k = cluster size.
Midcingulate gyrus 31 R 10 −40 30 5 3.43
Middle frontal gyrus 8 R 38 16 46 6 3.25

p < .001(uncorrected), k > 5, BA: Brodmann area. L/R:left/right, peak

voxel = MNI xyz co-ordinates, k = cluster size.

3.2. Overall neural patterns in each group

First, we examined BOLD responses to the contrast images (at-

tachment targets > control targets vs control targets > attachment
targets) in each group.

3.2.1. Prototypical secure group

The prototypical secure group showed activations in left inferior
frontal gyrus, dorsolateral prefrontal cortex, the caudate and putamen
in basal ganglia area, and the anterior cingulate gyrus and amygdala to
attachment targets compared to control targets; no activation was sig-
nificantly greater to control targets compared to attachment targets
(Table 3 and Fig. 2).

3.2.2. Secure but restricted group

The children who were assessed as the group of secure but restricted
showed no significant differences between the two conditions, in-
dicating that the neural activations of attachment targets and control
targets were processed similarly in this group.
3.2.3. Insecure group
The insecure group showed the opposite pattern of brain activation
compared the prototypical secure group. There were no significantly
greater activations to attachment than control targets, but the pre-
cuneus, middle temporal gyrus, and inferior temporal gyrus were sig-
nificantly more activated to control compared attachment targets
(Table 4 and Fig. 3).
3.3. Group differences
In the factorial model of conditions (2) and groups (3), we found the
Fig. 3. Brain activation in insecure group: peak activations in precuneus.
main effects of conditions and the interactions between condition and
group (pcorr < 0.05). The main effects of condition were found mostly
in superior parietal lobule, superior frontal gyrus, and precuneus. Post-
hoc analysis showed the attachment targets activated the superior
frontal gyrus, inferior frontal gyrus, and medial frontal gyrus compared
to control targets, but control targets were strongly related to the ac-
tivation of precuneus (Table 5 and Fig. 4).
The interaction of groups and conditions revealed significant
(pcorr < 0.05) activity in the basal ganglia including the globus pal-
lidus, putamen, subthalamic nucleus, and claustrum (Table 6 and
Fig. 5).
In a post-hoc analysis, the prototypical secure group children
showed greater activation (pcorr < 0.05) in regions of basal ganglia
during processing of the attachment targets compared to other two

Fig. 2. Brain activation in prototypical secure group: peak activations in inferior frontal gyrus (A), basal ganglia and striatum area (B), and anterior cingulate gyrus
and dorsolateral prefrontal cortex (C).

51
E.J. Choi et al. Brain and Cognition 124 (2018) 47–56

Table 5 Table 6
Main effects of condition. Interaction between conditions and groups.
Regions BA L/R Peak voxel (x,y,z) k z Regions BA L/R Peak voxel k z

Main effects of conditions Globus pallidus L −14 0 −10 53 5.60

Superior parietal lobule 7 L −26 −72 34 215 7.22 Putamen L −22 8 −14 4.68
Precuneus 7 L −22 −72 58 4.44 Subthalamic nucleus L −6 −12 −10 4.67
Dorsolateral prefrontal cortex 8 R 10 36 62 18 5.38 Globus pallidus R 10 0 −6 25 5.42
Precuneus 19 R 30 −76 38 15 4.75 Parahippocampal gyrus R 22 −4 −22 4.82
Precuneus 7 R 26 −68 30 4.42 Globus pallidus R 18 −4 −10 4.38
Precentral gyrus 6 L −38 4 30 10 4.40 Claustrum R 38 −8 6 23 4.93
Putamen R 22 −8 10 4.50
Post-hoc: Attachment > Control
Dorsolateral prefrontal cortex 8 R 10 36 62 25 5.50
Middle temporal gyrus 21 L −58 0 −22 6 4.90
FWE p < .05, k5, BA: Brodmann area. L/R:left/right, peak voxel = MNI xyz
Inferior frontal gyrus 47 L −34 20 −18 7 4.78 co-ordinates, k = cluster size.
Medial frontal gyrus 10 L −10 44 −10 7 4.59

Post-hoc: Control > Attachment

Precuneus 19 L −26 −72 34 247 7.30
Precuneus 7 L −22 −72 58 4.58
Precuneus 7 L −10 −76 50 4.22
Precuneus 19 R 30 −76 38 41 4.87
Precuneus 7 R 26 −68 30 4.56
Middle occipital gyrus 19 R 34 −80 22 4.26
Precental gyrus 6 L −38 4 30 17 4.55

FWE p < .05, k5, BA: Brodmann area. L/R:left/right, peak voxel = MNI xyz
co-ordinates, k = cluster size.

groups (Table 7 and Fig. 6).

We also examined the correlation between the basal ganglia activity
and attachment rating scores. The ROIs were created using peak MNI
coordinates which showed significant effects on group by condition
interaction. We did not find any correlation with children’s reports for
attachment security (ASS scores) in basal ganglia activity, but did find
significantly positive correlations with the emotional openness score,
one of the most important subscales in the Separation Anxiety Test
rating scales, in the left (x,y,z coordinates = −14, 0, 10; dia- Fig. 5. Interaction between condition and group: peak activations in basal
meter = 5 mm; r = 0.559; p < .008) and right globus pallidus (x,y,z ganglia.
coordinates = 10, 0, −6; diameter = 5 mm; r = 0.502; p < .020).
4.1. Frontal areas
4. Discussion
Inferior, medial and dorsolateral prefrontal areas were active during
In this study, we investigated the neural bases of attachment se- the processing of emotions in attachment-related scenes in children
curity in normally developing children by activating their attachment with prototypical secure attachment. These frontal cortical areas are
system in the fMRI environment. The results showed that children ex- related to modulating evaluative judgments (Cunningham, Johnson,
hibited different patterns of neural activity according to their level of Gatenby, Gore, & Banaji, 2003). The prefrontal contributions to affec-
attachment security in several frontal, limbic, and basal ganglia areas. tive evaluation are known for integrating information regarding

Fig. 4. Post-hoc analysis results for the main effect of condition brain activity of: (A) the medial prefrontal cortex in the attachment condition, (B) the precuneus in
the control condition.

52
E.J. Choi et al. Brain and Cognition 124 (2018) 47–56

Table 7 attachment-related scenes in this study depicted a boy, who was being
Comparisons between groups for brain activation in attachment condition. treated unfairly by his mother, and participants were asked to infer the
Regions BA L/R Peak voxel (x,y,z) k z boy’s emotional state. One of the most distinguishable features of the
prototypical securely attached children was a sense that they had al-
Prototypical secure > secure but restricted ready given thought to their feelings about parents and separation is-
globus pallidus R 10 0 −6 31 5.30
sues in general (Resnick, 1993). This did not seem to be a new topic to
amygdala R 22 −4 −22 4.96
putamen R 26 8 −10 52 5.11
them because they knew how they would feel and they had access to
claustrum R 38 −4 2 5.03 these feelings in a very consistent way. Another feature of this group
putamen R 26 −8 10 4.72 was that they displayed a strong use of self in their answers, that is, they
globus pallidus L −14 0 −6 13 4.82 could talk about feelings in an open and vulnerable way and related the
subthalamic nucleus L −6 −12 −10 5 4.65
pictured situation to their own experiences (i.e., “This boy feels sad
anterior cingulate gyrus 24 R 6 32 10 4.31
because he missed his parents… this happened to me and I was really
Prototypical secure > insecure
sad and lonely.”). Therefore, our findings in frontal areas suggest that
globus pallidus L −14 0 −10 92 7.68
globus pallidus R 10 0 −6 59 6.45 attachment security is functionally correlated with frontal areas which
claustrum R 38 −8 6 5.61 support the skill of understanding the mental states of others.
putamen R 26 8 −10 5.32
precuneus 31 R 26 −76 22 15 5.77
4.2. Limbic areas
cuneus 18 R 18 −80 18 5.46
cuneus 18 L −14 −88 18 9 5.76
amygdala R 22 −4 −22 5 4.40 The limbic system is well-known as the emotional network in our
brain and activation of these areas was not surprising as the task di-
FWE p < .05, k5, BA: Brodmann area. L/R:left/right, peak voxel = MNI xyz rectly contrasted two types of information, emotional or physical in-
co-ordinates, k = cluster size. formation. The anterior/midcingulate cortex and amygdala were acti-
vated during the processing of emotions in attachment-related scenes in
contextual variables, personal values and memories with information children with prototypical secure attachment.
provided by perceptual evaluative system. The medial prefrontal cortex Previous neuroimaging studies support the theory of dorsal-cogni-
has been found to be active in mentalizing tasks where participants tive vs. ventral-emotional functional dissociation within the anterior
infer the state of others (Frith & Frith, 2006), especially in making in- cingulate cortex (Bush, Luu, & Posner, 2000; Whalen et al., 1998), and
ferences about affective aspects (Völlm et al., 2006). especially the ventral ACC which was related to social feedback such as
Recently, attachment theory has expanded and connected with the being accepted or rejected (Somerville, Heatherton, & Kelley, 2006),
term “mentalization”. The ability to mentalize, the ability to reflect and social pain (Tracey & Mantyh, 2007; Vogt & Sikes, 2009). The ACC
upon and understand one’s state of mind and to have insight into what is specialized for affective processes, whereas the midcingulate cortex
one is feeling, is closely linked with attachment security. Fonagy and (MCC) is more specialized for cognitive processes. More recent imaging
Bateman (2005) proposed that caregivers’ insightful understanding of studies implicate the MCC in the regulation of autonomic activity and
children’s experience with appropriate feedback helps children learn the perception and production of emotion (Kober et al., 2008), and
how to understand what they are experiencing. This modeling ulti- especially the anterior subdivision of MCC was linked to the experience
mately culminates in children learning to reflect upon, and understand of more intense states of negative affect (Mechias, Etkin, & Kalisch,
their own states of mind. Importantly, this progression depends on a 2010; Sehlmeyer et al., 2009), delivery of pain (Farrell, Laird, & Egan,
healthy and consistent emotional interaction between children and 2005; Rainville, Duncan, Price, Carrier, & Bushnell, 1997) and other
caregivers, which supports the building of a secure attachment. How- aversive stimuli (Peyron, Laurent, & Garcia-Larrea, 2000; Vogt, 2005).
ever, when early caregivers are unable to reflect on children’s state of Amygdala activity is also critical for fear learning (LeDoux, 2003), but
mind, children do not receive the instruction they require, via caregiver there is substantial evidence that amygdala activity codes more gen-
modeling, to develop this important capacity. Therefore, their learning erally for relevant and arousing stimuli (Sanders, Wiltgen, & Fanselow,
how to understand their own thoughts and feelings or thoughts and 2003), and this may be especially true during childhood, when arousing
feeling of others is negatively impacted. stimuli, regardless of valence, effectively recruit amygdala activity
We found connections between attachment security and the ability (Heatherton & Wagner, 2011).
of mentalization during attachment interviews with children. The Emotional openness and regulation are linked to the quality of

Fig. 6. Comparison between groups of brain activation in the attachment condition (6A: Prototypical secure > secure but restricted; 6B: Prototypical secure >
insecure).

53
E.J. Choi et al.
attachment (Cassidy & Shaver, 2002). Being emotionally open and able
to verbalize feelings of vulnerability means that children can imagine
feelings of sadness, loneliness, anger and fear without experiencing
overwhelming tension or conflict. The secure style of attachment
manifests as open and flexible regulation whereas the avoidant style of
attachment tends to minimize affect and thus can be understood in
terms of the overregulation of affect. These problems in emotional
dysregulation can be tied to various types of pathology (Gergely et al.,
2002). In our study, securely attached children tended to be surprised
when the mother was dealing with her son unfairly, and they could also
access feelings of needing and missing those close to them. For example,
they verbalized some specific negative feelings clearly and sponta-
neously in a coherent manner (i.e., sad, angry, afraid, missing someone,
etc.). In contrast, vague feelings such as ‘not good’, ‘strange’, or the
failure to provide appropriate justifications are indicative of a lack of
emotional openness.
A recent neuroimaging study showed that during childhood and
adolescence, amygdala activity was preferentially recruited by images
of the child’s mother and this activity correlated with a greater ten-
dency to approach the mother, such as faster reaction times to maternal
stimuli during an in-scanner behavioural task or an out-of scanner
child-report of attachment related behaviour (Tottenham et al., 2012).
However it was not clear if amygdala activity was related to attachment
or just in response to a familiar face or emotional arousal. In this study
we determined that amygdala activity differed according to the level of
attachment security, and thus could represent the neurobiology of at-
tachment security.
4.3. Basal ganglia
The caudate and putamen are major components of the basal
ganglia, and make up the dorsal striatum which is highly related to
initiating motivationally relevant behaviours and reward processing.
The caudate is known as a key region to control approach-attachment
behaviours in animal models (Burkett, Spiegel, Inoue, Murphy, &
Young, 2011; Villablanca, 2010). These areas of the dopamine-rich
basal ganglia system are widely accepted to be associated with in-
itiating motivational behaviours necessary for survival such as food
(Knutson and Greer, 2008), but also are activated by strong social sti-
muli such as maternal and long-term intense romantic love (Acevedo,
Aron, Fisher, & Brown, 2012; Bartels & Zeki, 2004), which means that
social attachment in a close relationship would be considered as a
survival-related and highly-rewarding cue in our brain. The child-mo-
ther relationship, classically viewed as the first and strongest love
(Freud, 1949) and as the prototype of all later love relationships, not
only forms the matrix on which subsequent personality development
builds, but also provides the motivational core of social behaviour
throughout the lifespan.
Our findings showed that these areas were only activated by emo-
tional processing of attachment-related scenes in the prototypical se-
cure group. In the attachment interview, securely attached children had
a strong sense of motivation for being reunited with the attachment
figure and optimism for the outcome of the separation. That is, they
were obviously trying to regain the attachment figures or maintain
contact with them or seek comfort and support from them. They often
talked not only about their negative feelings during separation but also
that they would feel happy when reunited with their parents. This
finding can be interpreted that securely attached children responded to
attachment-related cues at an automatic processing level, as those cues
had been useful for their survival and rewarding. This was reflected at
the neural level by greater activation to attachment-related cues.
However, children who were lower on attachment security may lose
their biological motivation to attachment-related cues as those cues had
not been useful for their survival or not rewarding. They seem to have
developed a mechanism to not pay attention to these cues. This inter-
pretation is plausible as most of the insecurely attached children in our
54
Brain and Cognition 124 (2018) 47–56
study had avoidant attachment type and a key feature of avoidant at-
tachment is to minimize or ignore their vulnerability of being attached
to close others, and to overestimate their own ability to be independent.
When we directly compared the groups, basal ganglia was the most
important region to show the group differences according to the level of
attachment including globus pallidus, claustrum, subthalamic nucleus.
The globus pallidus and claustrum are implicated in initiating motiva-
tional behaviours and the computation of affective value (Arkadir,
Morris, Vaadia, & Bergman, 2004; Wager, Davidson, Hughes, Lindquist,
& Ochsner, 2008). Especially the striatum, comprising of globus pal-
lidus, putamen, and caudate nucleus, has been shown to mediate be-
tween socio-emotional information processing and attachment security
in adult studies (Schneider et al., 2012; Vrtička & Vuilleumier, 2012;
Vrticka et al., 2008; Vrticka, 2012). However there is little research
which demonstrates the role of the striatum as a functional basis of
attachment security in typically developing children. Our results sug-
gest that this pattern emerges in young childhood. Furthermore, we
found an association of the globus pallidus activity and children’s
emotional openness scores, during emotional processing in attachment-
related stimuli. Emotional openness, one of the most important sub-
scales in the Separation Anxiety Test ratings, represents the extent to
which the child is able to express feeling such as loneliness, sadness,
fear, or anger and is a key aspect of secure attachment. This finding
builds on the growing attention on the striatum as a neural basis of
secure attachment by showing for the first time the association of the
globus pallidus activity, attachment classification, and emotional
openness rating. Notably, we did not find correlations between brain
activity and children’s perception of attachment quality. There is evi-
dence that at the beginning of middle childhood insecure patterns be-
come sex biased, with boys showing more avoidance and girls showing
more anxiety (Del Giudice & Belsky, 2010). Thus, middle childhood
boys start to show traits such as aggression, independence and inflated
self-esteem and these traits may affect their perception of attachment
quality. Our data also showed that self-reported attachment security
from children was not significantly different among the groups even
though securely attached children tended to report higher attachment
security scores.
4.4. Precuneus
In this study, the precuneus was the region most implicated for
physical vs. emotional information processing in attachment-related
scenes; insecurely attached children showed precuneus activity only for
physical compared to emotional information processing. Recent neu-
roimaging findings in healthy subjects suggest a central role for the
precuneus in a wide range of highly integrated tasks, including visuo-
spatial imagery, episodic memory retrieval and self-processing opera-
tions, such as first-person perspective taking and experience of agency
(Cavanna & Trimble, 2006). Our control condition was designed to
control the attachment system by guiding the focus to physical aspects
of the stimuli, such as identifying the objects beside the child, the
spatial relationship of the objects, or the activities of the child. These
questions seemed to lead to an effective attentional shift to visuo-spatial
imagery. Importantly, only insecurely attached children showed the
strong activity in precuneus during the processing of physical in-
formation. Attachment-related scenes used in the attachment interview
contain negative and aversive emotions, as their roles typically activate
subjects’ attachment system. However, children in the avoidant at-
tachment group were generally unable to identify feelings, especially in
the more stressful pictures. They could provide understanding of what
was going on in the situation, hence the precuneus activation, but they
were not able to connect these situations with specific and open feel-
ings. There was a quality of “shutting-off” of attachment among these
children (Resnick, 1993). Some children in the avoidant attachment
group showed explicit refusal or resistance toward completing inter-
views or gave sparse or one-word answers. Some other of these children
E.J. Choi et al.
seemed to express some open feelings, but immediately followed by
discounting, devaluing or minimizing these feelings.
5. Conclusions
The present study determined the neural basis of attachment se-
curity in typically developing children. We identified the neural activity
when children’s attachment system was engaged in an fMRI environ-
ment. Our data support the idea that attachment security could be
detected at the neural level corresponding to the behavioural differ-
ences in the attachment interview. Securely and insecurely attached
children showed distinct patterns of neural activity in the frontal,
limbic and basal ganglia area which included the dorsolateral pre-
frontal cortex, amygdala, cingulate cortex and striatum. These regions
have a key role in socio-emotional information processing and also
represent a brain network related to approach and avoidance motiva-
tion in humans (Spielberg et al., 2012). Especially the striatum, strongly
linked to reward processing underpinning social approach and avoid-
ance motivation, showed the largest effects in these differences. The
model of attachment is understood as a motivational and behavioural
system that directs children to seek proximity with a familiar caregiver
when they are alarmed. Our results indicate that the quality of at-
tachment configures the approach and avoidance motivational system
in our brain, mediated by the striatum.
Acknowledgments
For generous support, the authors thank Brain Imaging Center of
department of Brain and Cognitive Sciences in Seoul National
University.
Financial disclosures
This research was supported by National Research Foundation of
Korea (NRF-2013S1A5A2A01019722/ NRF-2014S1A5B6038081).
References
Acevedo, B. P., Aron, A., Fisher, H. E., & Brown, L. L. (2012). Neural correlates of marital
satisfactions and well-being: Reward, empathy, and affect. Clinical
Neuropsychiatry, 9(1).
Achenbach, T. M., & Rescorla, L. A. (2001). Manual for the ASEBA School-Age Forms &
Profiles. Burlington, VT: University of Vermont, Research Center for Children, Youth,
& Families.
Ainsworth, M. (1973). The development of infant-mother attachment [w:] B. Caldwell,
HN Ricciuti (red.). Review of Child Development Research, 1–94.
Ainsworth, M. D. S., Blehar, M. C., Waters, E., & Wall, S. N. (2015). Patterns of attach-
ment: A psychological study of the strange situation. Psychology Press.
Arkadir, D., Morris, G., Vaadia, E., & Bergman, H. (2004). Independent coding of
movement direction and reward prediction by single pallidal neurons. Journal of
Neuroscience, 24(45), 10047–10056.
Aron, A., Fisher, H., Mashek, D. J., Strong, G., Li, H., & Brown, L. L. (2005). Reward,
motivation, and emotion systems associated with early-stage intense romantic love.
Journal of Neurophysiology, 94(1), 327–337.
Arsalidou, M., Duerden, E. G., & Taylor, M. J. (2013). The centre of the brain:
Topographical model of functions of the basal ganglia. Human Brain Mapping, 34,
3031–3054.
Bartels, A., & Zeki, S. (2004). The neural correlates of maternal and romantic love.
Neuroimage, 21(3), 1155–1166.
Benoit, D., & Parker, K. C. (1994). Stability and transmission of attachment across three
generations. Child Development, 65(5), 1444–1456.
Bowlby, J. (1982). Attachment and loss: Retrospect and prospect. American Journal of
Orthopsychiatry, 52(4), 664.
Bowlby, J. (1969). Attachment and loss, Vol. 1 Loss. New York: Basic Books.
Bowlby, J. (1973). Attachment and loss, Vol. 2: Separation. New York: Basic Books.
Bretherton, I., & Munholland, K. A. (2008). Internal working models in attachment re-
lationships: Elaborating a central construct in attachment theory. In I. J. Cassidy & P.
R. Shaver (Red.), Handbook of attachment: Theory, research and clinical applications
(s. 102–127).
Buchheim, A., Erk, S., George, C., Kächele, H., Kircher, T., Martius, P., ... Walter, H.
(2008). Neural correlates of attachment trauma in borderline personality disorder: A
functional magnetic resonance imaging study. Psychiatry Research: Neuroimaging,
163(3), 223–235.
Buchheim, A., Erk, S., George, C., Kächele, H., Ruchsow, M., Spitzer, M., ... Walter, H.
55
Brain and Cognition 124 (2018) 47–56
(2006). Measuring attachment representation in an fMRI environment: A pilot study.
Psychopathology, 39(3), 144–152.
Burkett, J. P., Spiegel, L. L., Inoue, K., Murphy, A. Z., & Young, L. J. (2011). Activation of
μ-opioid receptors in the dorsal striatum is necessary for adult social attachment in
monogamous prairie voles. Neuropsychopharmacology, 36(11), 2200.
Bush, G., Luu, P., & Posner, M. I. (2000). Cognitive and emotional influences in anterior
cingulate cortex. Trends in Cognitive Sciences, 4(6), 215–222.
Cassidy, J., & Shaver, P. R. (Eds.). (2002). Handbook of attachment: Theory, research, and
clinical applications. Rough Guides.
Cavanna, A. E., & Trimble, M. R. (2006). The precuneus: A review of its functional
anatomy and behavioural correlates. Brain, 129(3), 564–583.
Cohen, M., & Shaver, P. (2004). Avoidant attachment and hemispheric lateralisation of
the processing of attachment-and emotion-related words. Cognition and Emotion,
18(6), 799–813.
Cunningham, W. A., Johnson, M. K., Gatenby, J. C., Gore, J. C., & Banaji, M. R. (2003).
Neural components of social evaluation. Journal of Personality and Social Psychology,
85(4), 639.
Debbané, M., Badoud, D., Sander, D., Eliez, S., Luyten, P., & Vrtička, P. (2017). Brain
activity underlying negative self-and other-perception in adolescents: The role of
attachment-derived self-representations. Cognitive, Affective, & Behavioral
Neuroscience, 17(3), 554–576.
Del Giudice, M., & Belsky, J. (2010). Sex differences in attachment emerge in middle
childhood: An evolutionary hypothesis. Child Development Perspectives, 4(2), 97–105.
Diamond, N., & Marrone, M. (2003). Attachment and intersubjectivity. Whurr Publishers.
Donges, U. S., Kugel, H., Stuhrmann, A., Grotegerd, D., Redlich, R., Lichev, V., ...
Dannlowski, U. (2012). Adult attachment anxiety is associated with enhanced auto-
matic neural response to positive facial expression. Neuroscience, 220, 149–157.
Farrell, M. J., Laird, A. R., & Egan, G. F. (2005). Brain activity associated with painfully
hot stimuli applied to the upper limb: A meta-analysis. Human Brain Mapping, 25(1),
129–139.
Fisher, H., Aron, A., & Brown, L. L. (2005). Romantic love: An fMRI study of a neural
mechanism for mate choice. Journal of Comparative Neurology, 493(1), 58–62.
Fonagy, P., & Bateman, A. (2005). Attachment theory and mentalization-oriented model
of borderline personality disorder. The American Psychiatric Publishing Textbook of
Personality Disorders, 187–207.
Freud, S. (1949). An outline of psycho-analysis (J. Strachey, Ed. & Trans.). The standard
edition of the complete psychological works of Sigmund Freud, 23.
Friston, K. J., Holmes, A. P., Price, C. J., Büchel, C., & Worsley, K. J. (1999). Multisubject
fMRI studies and conjunction analyses. Neuroimage, 10(4), 385–396.
Frith, C. D., & Frith, U. (2006). The neural basis of mentalizing. Neuron, 50(4), 531–534.
Gergely, G., Fonagy, P., Jurist, E., & Target, M. (2002). Affect regulation, mentalization,
and the development of the self.
Gwak, G., Oh, S., & Kim, C. (2011). K-WISC-IV (Korean Wechsler Intelligence Scale for
Children-IV). Manual for experts. Seoul: Hakjisa.
Hansburg, H. G. (1972). Adolescent separation anxiety: A method for the study of adolescent
separation problems. Springfield, IL: Thomas.
Heatherton, T. F., & Wagner, D. D. (2011). Cognitive neuroscience of self-regulation
failure. Trends in Cognitive Sciences, 15(3), 132–139.
Kerns, K. A., Aspelmeier, J. E., Gentzler, A. L., & Grabill, C. M. (2001). Parent-child at-
tachment and monitoring in middle childhood. Journal of Family Psychology, 15(1),
69–81.
Knutson, B., & Greer, S. M. (2008). Anticipatory affect: Neural correlates and con-
sequences for choice. Philosophical Transactions of the Royal Society B: Biological
Sciences, 363(1511), 3771–3786.
Kober, H., Barrett, L. F., Joseph, J., Bliss-Moreau, E., Lindquist, K., & Wager, T. D. (2008).
Functional grouping and cortical–subcortical interactions in emotion: A meta-ana-
lysis of neuroimaging studies. Neuroimage, 42(2), 998–1031.
LeDoux, J. (2003). The emotional brain, fear, and the amygdala. Cellular and Molecular
Neurobiology, 23(4–5), 727–738.
Lenzi, D., Trentini, C., Pantano, P., Macaluso, E., Iacoboni, M., Lenzi, G. L., & Ammaniti,
M. (2008). Neural basis of maternal communication and emotional expression pro-
cessing during infant preverbal stage. Cerebral Cortex, 19(5), 1124–1133.
Lemche, E., Giampietro, V. P., Surguladze, S. A., Amaro, E. J., Andrew, C. M., Williams, S.
C., ... Simmons, A. (2006). Human attachment security is mediated by the amygdala:
Evidence from combined fMRI and psychophysiological measures. Human Brain
Mapping, 27(8), 623–635.
Liu, Y., Ding, Y., Lu, L., & Chen, X. (2017). Attention bias of avoidant individuals to
attachment emotion pictures. Scientific Reports, 7, 41631.
Mechias, M. L., Etkin, A., & Kalisch, R. (2010). A meta-analysis of instructed fear studies:
Implications for conscious appraisal of threat. Neuroimage, 49(2), 1760–1768.
Mikulincer, M., & Shaver, P. R. (2007). Attachment in adulthood: Structure, dynamics, and
change. Guilford Press.
Minagawa-Kawai, Y., Matsuoka, S., Dan, I., Naoi, N., Nakamura, K., & Kojima, S. (2008).
Prefrontal activation associated with social attachment: Facial-emotion recognition
in mothers and infants. Cerebral Cortex, 19(2), 284–292.
Moutsiana, C., Fearon, P., Murray, L., Cooper, P., Goodyer, I., Johnstone, T., & Halligan,
S. (2014). Making an effort to feel positive: Insecure attachment in infancy predicts
the neural underpinnings of emotion regulation in adulthood. Journal of Child
Psychology and Psychiatry, 55(9), 999–1008.
Nitschke, J. B., Nelson, E. E., Rusch, B. D., Fox, A. S., Oakes, T. R., & Davidson, R. J.
(2004). Orbitofrontal cortex tracks positive mood in mothers viewing pictures of their
newborn infants. Neuroimage, 21(2), 583–592.
Norman, L., Lawrence, N., Iles, A., Benattayallah, A., & Karl, A. (2014). Attachment-se-
curity priming attenuates amygdala activation to social and linguistic threat. Social
Cognitive and Affective Neuroscience, 10(6), 832–839.
Peyron, R., Laurent, B., & Garcia-Larrea, L. (2000). Functional imaging of brain responses
E.J. Choi et al.
to pain. A review and meta-analysis (2000). Neurophysiologie Clinique/Clinical
Neurophysiology, 30(5), 263–288.
Rainville, P., Duncan, G. H., Price, D. D., Carrier, B., & Bushnell, M. C. (1997). Pain affect
encoded in human anterior cingulate but not somatosensory cortex. Science,
277(5328), 968–971.
Ranote, S., Elliott, R., Abel, K. M., Mitchell, R., Deakin, J. F. W., & Appleby, L. (2004). The
neural basis of maternal responsiveness to infants: An fMRI study. Neuroreport,
15(11), 1825–1829.
Redlich, R., Grotegerd, D., Opel, N., Kaufmann, C., Zwitserlood, P., Kugel, H., ...
Dannlowski, U. (2015). Are you gonna leave me? Separation anxiety is associated
with increased amygdala responsiveness and volume. Social Cognitive and Affective
Neuroscience, 10(2), 278–284.
Resnick, G. (1993). Measuring attachment in early adolescence: A manual for the adminis-
tration, coding and interpretation of the SAT for 11 to 14 year olds. Rockville, MD:
Westat.
Sander, K., Frome, Y., & Scheich, H. (2007). FMRI activations of amygdala, cingulate
cortex, and auditory cortex by infant laughing and crying. Human Brain Mapping,
28(10), 1007–1022.
Sanders, M. J., Wiltgen, B. J., & Fanselow, M. S. (2003). The place of the hippocampus in
fear conditioning. European Journal of Pharmacology, 463(1), 217–223.
Schneider, S., Brassen, S., Bromberg, U., Banaschewski, T., Conrod, P., Flor, H., ... Nees, F.
(2012). Maternal interpersonal affiliation is associated with adolescents' brain
structure and reward processing. Translational Psychiatry, 2(11), e182.
Sehlmeyer, C., Schöning, S., Zwitserlood, P., Pfleiderer, B., Kircher, T., Arolt, V., &
Konrad, C. (2009). Human fear conditioning and extinction in neuroimaging: A
systematic review. PloS one, 4(6), e5865.
Somerville, L. H., Heatherton, T. F., & Kelley, W. M. (2006). Anterior cingulate cortex
responds differentially to expectancy violation and social rejection. Nature
Neuroscience, 9(8), 1007.
Spielberg, J. M., Miller, G. A., Warren, S. L., Engels, A. S., Crocker, L. D., Banich, M. T., ...
Heller, W. (2012). A brain network instantiating approach and avoidance motivation.
Psychophysiology, 49(9), 1200–1214.
Takiguchi, S., Fujisawa, T. X., Mizushima, S., Saito, D. N., Okamoto, Y., Shimada, K., ...
Hiratani, M. (2015). Ventral striatum dysfunction in children and adolescents with
reactive attachment disorder: Functional MRI study. British Journal of Psychiatry
56
Brain and Cognition 124 (2018) 47–56
Open, 1(2), 121–128.
Tottenham, N., Shapiro, M., Telzer, E. H., & Humphreys, K. L. (2012). Amygdala response
to mother. Developmental Science, 15(3), 307–319.
Tracey, I., & Mantyh, P. W. (2007). The cerebral signature for pain perception and its
modulation. Neuron, 55(3), 377–391.
Villablanca, J. R. (2010). Why do we have a caudate nucleus. Acta Neurobiologiae
Experimentalis (Wars), 70(1), 95–105.
Vogt, B. A. (2005). Pain and emotion interactions in subregions of the cingulate gyrus.
Nature Reviews. Neuroscience, 6(7), 533.
Vogt, B. A., & Sikes, R. W. (2009). Cingulate nociceptive circuitry and roles in pain
processing: The cingulate premotor pain model. Cingulate Neurobiology and Disease,
312–339.
Völlm, B. A., Taylor, A. N., Richardson, P., Corcoran, R., Stirling, J., McKie, S., ... Elliott,
R. (2006). Neuronal correlates of theory of mind and empathy: A functional magnetic
resonance imaging study in a nonverbal task. Neuroimage, 29(1), 90–98.
Vrticka, P. (2012). Interpersonal closeness and social reward processing. Journal of
Neuroscience, 32(37), 12649–12650.
Vrtička, P., & Vuilleumier, P. (2012). Neuroscience of human social interactions and adult
attachment style. Frontiers in Human Neuroscience, 6.
Vrticka, P., Andersson, F., Grandjean, D., Sander, D., & Vuilleumier, P. (2008). Individual
attachment style modulates human amygdala and striatum activation during social
appraisal. PLoS One, 3, 11.
Vrticka, P., Lordier, L., Bediou, B., & Sander, D. (2014). Human amygdala response to
dynamic facial expressions of positive and negative surprise. Emotion, 14(1), 161.
Wager, T. D., Davidson, M. L., Hughes, B. L., Lindquist, M. A., & Ochsner, K. N. (2008).
Prefrontal-subcortical pathways mediating successful emotion regulation. Neuron,
59(6), 1037–1050.
Waters, E., Merrick, S., Treboux, D., Crowell, J., & Albersheim, L. (2000). Attachment
security in infancy and early adulthood: A twenty-year longitudinal study. Child
Development, 71(3), 684–689.
Whalen, P. J., Bush, G., McNally, R. J., Wilhelm, S., McInerney, S. C., Jenike, M. A., &
Rauch, S. L. (1998). The emotional counting Stroop paradigm: A functional magnetic
resonance imaging probe of the anterior cingulate affective division. Biological
Psychiatry, 44(12), 1219–1228.