Neonatal amygdala resting-state functional connectivity and socio-emotional development in very preterm children

Abstract

Very preterm children are more likely to exhibit difficulties in socio-emotional processing than their term-born peers. Emerging socio-emotional problems may be partly due to alterations in limbic system development associated with infants' early transition to extrauterine life. The amygdala is a key structure in this system and plays a critical role in various aspects of socio-emotional development, including emotion regulation. The current study tested the hypothesis that amygdala resting-state functional connectivity at term-equivalent age would be associated with socio-emotional outcomes in childhood. Participants were 129 very preterm infants (<33 weeks' gestation) who underwent resting-state functional MRI at term and received a neurodevelopmental assessment at 4-7 years (median = 4.64). Using the left and right amygdalae as seed regions, we investigated associations between whole-brain seed-based functional connectivity and three socio-emotional outcome factors which were derived using exploratory factor analysis (Emotion Moderation, Social Function and Empathy), controlling for sex, neonatal sickness, post-menstrual age at scan and social risk. Childhood Emotion Moderation scores were significantly associated with neonatal resting-state functional connectivity of the right amygdala with right parahippocampal gyrus and right middle occipital gyrus, as well as with functional connectivity of the left amygdala with the right thalamus. No significant associations were found between amygdalar resting-state functional connectivity and either Social Function or Empathy scores. The current findings show that amygdalar functional connectivity assessed at term is associated with later socio-emotional outcomes in very preterm children.

Keywords: child development; resting-state fMRI; socio-emotional outcomes; very preterm.

Graphical abstract

Figure 1

Voxel-wise statistical maps and regression partial plots depicting associations between amygdalar rs-FC and Emotion Moderation scores. Whole-brain voxel-wise statistical maps are family-wise error corrected. Right amygdala (rAMYG)–right middle occipital gyrus (rMOG): A. statistical map of rMOG cluster; B. association between rAMYG–rMOG and Emotion Moderation score. rAMYG–left middle occipital gyrus (lMOG): C. statistical map of lMOG cluster; D. association between rAMYG–lMOG and Emotion Moderation score. rAMYG–right parahippocampal gyrus (rPHG): E. statistical map of rPHG cluster; F. association between rAMYG–rPHG and Emotion Moderation score. Left amygdala (lAMYG)–right thalamus (rTHAL): G. statistical map of rTHAL cluster; H: association between lAMYG–rTHAL and Emotion Moderation score. All regression partial plots were created after outlier deletion. Yellow, positive associations; blue, negative associations; R, right; L, left. As images are not in MNI template space but rather in a neonatal template space, we have opted to use a crosshair to indicate exact peak position and the AAL labels to describe these regions.