Cortical Structure in Nodes of the Default Mode Network Estimates General Intelligence

Abstract

Introduction

A growing number of studies implicate functional brain networks in intelligence, but it is unclear if network nodal structure relates to intelligence.

Methods

Using MRI, we studied the relationship of the general intelligence factor (g) with cortical thickness (CT), local gyrification index (LGI), and voxel-based morphometry in the nodes of the default mode network (DMN) and task-positive network (TPN) in a cohort of 44 young, healthy adults. Employing a novel strategy, we performed repeated analyses with multiple sets of g estimates to remove false positives.

Results

CT and LGI in medial and temporal nodes of the DMN were reliably correlated with g (p < 0.05; Pearson's coefficient: ‑0.52 to ‑0.25 and 0.22 to 0.41, respectively). Linear regression models were developed with these parameters to estimate individual g scores, with a median adj. R² of 0.25.

Conclusion

Cortical thickness and gyrification in key nodes of the Default Mode Network correlate with intelligence. Linear regression models with these cortical parameters may provide an estimate of the g factor.