Sex-Biased Trajectories of Amygdalo-Hippocampal Morphology Change Over Human Development

Abstract

The amygdala and hippocampus are two adjacent allocortical structures implicated in sex-biased and developmentally-emergent psychopathology. However, the spatiotemporal dynamics of amygdalo-hippocampal development remain poorly understood in healthy humans.The current study defined trajectories of volume and shape change for the amygdala and hippocampus by applying a multi-atlas segmentation pipeline (MAGeT-Brain) and semiparametric mixed-effects spline modeling to 1,529 longitudinally-acquired structural MRI brain scans from a large, single-center cohort of 792 youth (403 males, 389 females) between the ages of 5 and 25 years old. We found that amygdala and hippocampus volumes both follow curvilinear and sexually dimorphic growth trajectories. These sex-biases were particularly striking in the amygdala: males showed a significantly later and slower adolescent deceleration in volume expansion (at age 20 years) than females (age 13 years). Shape analysis localized significant hot-spots of sex-biased anatomical development in sub-regional territories overlying rostral and caudal extremes of the CA1/2 in the hippocampus, and the centromedial nuclear group of the amygdala. In both sexes, principal components analysis revealed close integration of amygdala and hippocampus shape change along two main topographically-organized axeslow vs. high areal expansion, and early vs. late growth deceleration. These results bring greater resolution to our spatiotemporal understanding of amygdalo-hippocampal development in healthy males and females and discover focal sex-differences in the structural maturation of the brain components that may contribute to differences in behavior and psychopathology that emerge during adolescence.al., 2014) to a large, single-center longitudinal neuroimaging study of human brain development (Giedd et al., 2015), and (ii) estimating developmental trajectories with semi-parametric splinebased methods, which offer several advantages over classical polynomial approaches to curvefitting (Mills & Tamnes, 2014). We combine these methodological advances to capture curvilinear trajectories of amygdala and hippocampal volume across the transition from childhood and early adulthood in each sex, and pinpoint key maturational markers (Fjell et al., 2010; Raznahan, Shaw, et al., 2014; Shaw et al., 2008) that have remained unresolved for the amygdala and hippocampus: age of fastest volume change, age of greatest growth deceleration, and age at attainment of peak volume.Second, we move beyond analysis of bulk volume by modelling trajectories of surface area change at 5,245 points (vertices) across the amygdala and hippocampus to provide reference, surface-based "maturation maps" for these structures that are in line with those that are already available for the cortex, striatum, pallidum, and thalamus (Raznahan, Shaw, et al.,