BackHome
Refresh: OffView Personal Schedule
Individual Submission Summary
Share...

Direct link:

Cortical Metrics in Pre-Adolescence: A Latent Variable Analysis of sMRI Data in the ABCD Study

Wed, April 7, 11:45am to 12:45pm EDT (11:45am to 12:45pm EDT), Virtual

Abstract

Background: The brain undergoes rapid structural changes during adolescence, including reductions in cortical volume and thickness (i.e., cortical thinning; Tamnes et al., 2017), particularly in the frontal and parietal lobes (Bramen, 2012). These changes may be impacted by biological sex (De Bellis et al., 2001) and are likely associated with positive outcomes such as enhanced cognitive abilities (Squeglia et al, 2013). However, previous studies on cortical development in adolescence have been limited by small samples. Further, the literature lacks a comprehensive understanding of the factors that influence morphometric changes within specific cortical regions of interest (ROIs). Here, we aimed to examine the effects of age and sex on the volume and thickness of ROIs in a large, nationally representative sample of pre-adolescent youth from the Adolescent Brain Cognitive Development (ABCD) Study.

Methods: Our sample was comprised of 10,599 youth ages 9-10 (48% female; mean age=119 months, sd=7.5; DOI=10.15154/1519579). Participants were excluded for missing or poor quality structural magnetic resonance imaging (sMRI) data. Cortical volume and thickness were measured within all 68 ROIs of the Desikan atlas (Desikan et al., 2006), corrected for whole-brain cortical volume and thickness. To reduce dimensionality in the set of ROIs, we conducted principal component analyses (PCAs; using orthogonal rotation) with the volume and thickness of each region. Components that explained more than three percent of variance were extracted, as this resulted in representative yet parsimonious sets of components. Next, we used linear mixed effects models (LMEs) to assess for age- and sex-related effects on the component scores. Fixed effects included age in months, sex at birth, and an interaction between age and sex. Random effects covariates were study site and family (i.e., having a sibling in the study). Bonferroni correction was used to control for the number of LMEs (p=0.05/9=0.006).

Results: Three rotated components were extracted for thickness and six were extracted for volume (see Table 1 for variable loadings). Both sets of components explained 24% of the total variance in the respective cortical metrics. The LMEs revealed several significant age effects. Older children demonstrated greater thickness in a component comprised of temporal regions (T-RC3: t=3.91, p<0.001, squared semi-partial correlation [sr2]=0.001) and less thickness in components largely comprised of insular, orbitofrontal, cingulate, and occipital regions (T-RC1: t=-3.15, p=0.002, sr2=0.001; T-RC2: t=-5.55, p<0.001, sr2=0.002). Older children had greater volumes in a component comprised of orbitofrontal regions (V-RC2: t=3.42, p<0.001, sr2=0.001) and lower volumes in components largely comprised of parietal and occipital regions (V-RC1: t=-2.79, p=0.005, sr2=0.001; V-RC4: t=4.18, p<0.001, sr2=0.002). No significant sex or interaction effects were observed (p's>0.03).

Conclusions: Our findings suggest that age may have differential effects on the volume and thickness of cortical regions in pre-adolescence, indicating that cortical thinning may be normative for some ROIs but not others. However, the effect sizes of these results were small, likely due to the truncated age range of 9-10. Therefore, future studies should explore these relationships longitudinally as well as examine other factors that may influence morphometric changes during adolescence (e.g., SES and pubertal status).

Authors