Amygdala volumes and associations with socio-emotional competencies in preterm youth: cross-sectional and longitudinal data

Abstract

Background: Socio-emotional difficulties often result from very preterm (VPT) birth. The amygdala's developmental trajectory, including its nuclei, has been recognized as a significant factor in observed difficulties. This study aims to assess the relationship between amygdala volume and socio-emotional competencies in VPT children and adolescents.

Methods: Socio-emotional competencies were assessed, and amygdala volumes, including subnuclei, were extracted automatically from structural scans in a cross-sectional cohort of VPT (n = 75) and full-term (FT, n = 41) aged 6-14 years. Group differences in amygdala volumes were assessed using ANCOVA, and associations with socio-emotional competencies were studied using partial least squares correlation (PLSC). In a VPT subgroup, additional longitudinal data with amygdala volumes at term-equivalent age (TEA) were manually extracted, growth rates calculated, and associations with school-age socio-emotional competencies investigated using PLSC.

Results: Using cross-sectional data at school-age, amygdala volumes displayed comparable developmental patterns between the VPT and the FT groups. Greater volumes were associated with more emotional regulation difficulties in VPT and lower affect recognition competencies in FT. In the longitudinal VPT subgroup, no significant associations were found between amygdala volume trajectory and socio-emotional competencies.

Conclusion: Although our findings suggest typical amygdala development after VPT birth, further research is necessary to elucidate the developmental trajectory of amygdala and the role of resilience factors.

Impact: In our cohort, amygdala volumes, including subnuclei, displayed comparable developmental trajectories between the very preterm and the full-term groups. Higher amygdala volumes at school-age were associated with higher emotional regulation difficulties in the very-preterm born group, and with lower affect recognition abilities in full-term born children and adolescents. In a subgroup of very-preterm children and adolescents followed from birth to school-age, no significant associations were found between amygdala volumes at term-equivalent age and socio-emotional competencies at school-age.

Fig. 1. Illustration of amygdala subfield segmentation by FreeSurfer (version 7.1.1).

Amygdala subfields are shown from left to right in sagittal, coronal and axial views, respectively.

Fig. 2. Illustration of the right amygdala manual delineation using ITK-SNAP software, (version 3.6.0).

The right amygdala, in green, is shown from left to right in sagittal, coronal and axial views, respectively.

Fig. 3. Associations between amygdala volumetric measures and socio-emotional competencies at school-age in the very preterm (VPT) and full-term (FT) groups based on the PLSC analysis.

Each latent component is composed of a set of socio-emotional weights (different for each group) and amygdala volume weights (common to all participants), which indicate how strongly each socio-emotional and amygdala volume variable contribute to the multivariate socio-emotional/amygdala volume association. a The diverging graph shows socio-emotional bootstrap ratio z-scores for the VPT and the FT groups in dark blue and in light blue, respectively. Socio-emotional measures with an absolute bootstrap ratio z-score ≥ 3 or ≤−3 yield a robust contribution to the component and are marked with a star. b The diverging graph shows amygdala volume bootstrap ratio z-scores. Amygdala volumes with an absolute bootstrap ratio z-score ≥ 3 or ≤−3 yield a robust contribution to the component and are marked with a star.

Fig. 4. Line plot showing individual longitudinal growth of the total amygdala volume from TEA to school-age.

x-axis: age in months; y-axis: total amygdala volume in mm3. Line colors indicate amygdala growth rates relative to the amygdala volume at TEA (faster and slower growth are represented by red and blue colors, respectively).