Resting State fMRI and Attachment Difficulties in Children

Background: Autism spectrum disorder (ASD) can be a risk factor for developing secure attachment. Correspondingly, attachment problems also disturb social development in the typically developing population, and in severe cases, attachment deprivation in a critical period can lead to autistic symptoms, having life-long effects on the individual’s mental health. Although the etiology of ASD is more strongly associated with genetics, children with ASD and children with attachment problems may present with some overlapping difficulties in building social relationships. Despite the extensive neuroimaging studies on neural correlates underlying ASD, little is known about the neural correlates of attachment disturbance.

Objectives: In the present study, we compared DMN functional connectivity between secure and insecure attachment in typically developing children and correlated connectivity strength with neuropsychological measures to determine if attachment problems affect DMN function.

Methods: Resting-state fMRI data on a Siemens 3T MRI using 32 channel head coil was collected in typically developing 9-10-year-old boys (N=80, Mean age/SD = 9.62/0.69). Of these children, 46 were assessed as secure attachment and 34 as insecure according to the semi-structured attachment interview, The Separation Anxiety Test (Resnick, 1993). Subject’s data were slice-time and motion corrected, smoothed, intensity normalized. Signal from the white matter and CSF were regressed out. The default-mode network (DMN) connectivity, which represents a relaxed mental state, stimulus-independent thoughts or self-referential processing, was investigated between secure and insecure attachment children using seed-based analysis for five core DMN seeds (pC, precuneus; PCC, posterior cingulate cortex; vACC, ventral anterior cingulate cortex; mPFC, medical prefrontal cortex; vmPFC, ventromedial prefrontal cortex), using FSL’s FEAT program. Statistical images were thresholded using clusters determined by Z >2.3 and a corrected cluster significance threshold of p = 0.05.

Results: We found significantly increased functional connectivity among core regions of DMN in children with insecure attachment such that the vACC, vmPFC and PCC. Increased vmPFC connectivity was also found in dorsomedial prefrontal and hippocampal regions, which are known as parts of two subsystems of DMN, in children with insecure attachment. In addition, the increased functional connectivity in DMN was negatively correlated with the children’s emotional openness scores (vACC connectivity with PCC: r=-0.44, p=0.00).

Conclusions: Our results are the first to show that early attachment experience contributes to the shaping of core DMN functional connectivity in typically developing children, implicating the importance of early experience. Given that atypical DMN functional connectivity has often been reported in ASD, our results may offer one model as to why some behavioural manifestations may overlap between these two conditions.