The Genetic Contributions to Maturational Coupling in the Human Cerebrum: A Longitudinal Pediatric Twin Imaging Study

Abstract

Although prior studies have demonstrated that genetic factors play the dominant role in the patterning of the pediatric brain, it remains unclear how these patterns change over time. Using 1748 longitudinal anatomic MRI scans from 792 healthy twins and siblings, we quantified how genetically mediated inter-regional associations change over time via multivariate longitudinal structural equation modeling. These analyses found that genetic correlations for both lobar volumes and cortical thickness are dynamic, with relatively static effects on surface area. While genetic correlations for lobar volumes decrease over childhood and adolescence, in general they increase for cortical thickness in the second decade of life. Quantification of how genetic factors influence maturational coupling improves our understanding of typical neurodevelopment and informs future molecular genetic analyses.

Figure 1.

Statistical significance of genetically mediated maturational coupling for cerebral volumes, cortical thickness, and surface area. The heatmaps display −log₁₀P-values for all pairwise combinations of ROIs; the top row shows statistical significance of absolute genetic effects on pairwise genetic covariance, while the bottom row shows statistically significant changes in genetic covariance over childhood and adolescence. On the margins, dendograms from hierarchical cluster analyses are provided.

Figure 2.

Age-related changes in genetic correlations (r_G) for cerebral volumes. Panel A displays changes in genetic correlations for all ROI pairs, with 95% confidence intervals shown in gray. Asterisks identify ROI pairs with significant age-related changes after FDR-correction. The heatmap (B) shows the same trajectories sorted via hierarchical cluster analysis. The numbers on the right correspond to the numbers in each panel of A. A dynamic version of these results available in Supplementary Movie 1.

Figure 3.

Age-related changes in r_G for cortical thickness with 95% confidence intervals shown in panel A. Asterisks identify ROI pairs with significant age-related changes after FDR-correction. A heatmap of the same results are shown in Panel B. Supplemental Movie 2 presents the same data dynamically.