611 - Neonatal amygdala microstructure and structural connectivity are associated with autistic traits at 2 years of age

Poster Number: 611
Publication Number: 611.1384

Background: Prenatal exposure to maternal stress is associated with behavioural and mental health problems later in life. Transmission of prenatal stress to foetal neurodevelopment appears to occur via hypothalamic-pituitary-adrenal axis activity on the developing amygdala. Previously, we reported that maternal hair cortisol levels in pregnancy are associated with altered neonatal amygdala microstructure and structural connectivity, but behavioural correlates of these changes in the amygdala are unknown.

Objective: To investigate associations of neonatal amygdala microstructure and structural connectivity with neurodevelopmental outcome at 2 years of age.

Design/Methods: We included a subset of Theirworld Edinburgh Birth Cohort participants with data from brain MRI at term-equivalent age and any developmental assessment at 2 years corrected age (Tables 1 and 2). Diffusion tensor imaging and neurite orientation dispersion and density imaging metrics were calculated for left and right amygdalae. We used anatomically constrained tractography to obtain fractional anisotropy-weighted structural connectivity from left and right amygdalae to 6 ipsilateral regions of interest: insula, putamen, thalamus, inferior temporal gyrus, medial orbitofrontal cortex, rostral anterior cingulate cortex. Developmental assessment measures were Q-CHAT score (autistic traits), ECBQ factor scores (temperament), BRIEF-P global score (executive function) and, for the preterm group only, Bayley-III cognitive, language, social-emotional and adaptive behaviour domain scores (Table 2). We used linear regression to model amygdala-outcome associations, adjusting for gestational age at birth and at scan, sex, maternal education, and risk of maternal postnatal depression. We corrected for multiple comparisons using the false discovery rate separately within each outcome.

Results: Increased amygdala mean diffusivity bilaterally associated with lower Q-CHAT scores (left: β = -0.32, 95% CI [-0.54 to -0.09]; right: β = -0.38, 95% CI [-0.59 to -0.16]). Increased left amygdala neurite density index (β = 0.36, 95% CI [0.12 to 0.60]) and left amygdala-putamen connectivity (β = 0.32, 95% CI [0.10 to 0.53]) associated with higher Q-CHAT scores. Other neurodevelopmental outcomes did not significantly associate with amygdala microstructure or connectivity (Figure 1).

Conclusion(s): Our results suggest that microstructural variation in the neonatal amygdala may be important in the development of autistic traits.