The heritability of multi-modal connectivity in human brain activity.

Giles L Colclough,Stamatios N Sotiropoulos,Mark W Woolrich,Anderson M Winkler,Matthew F Glasser,David C Van Essen,Stephen M Smith,Thomas E Nichols

doi:10.7554/elife.20178

Abstract

Patterns of intrinsic human brain activity exhibit a profile of functional connectivity that is associated with behaviour and cognitive performance, and deteriorates with disease. This paper investigates the relative importance of genetic factors and the common environment between twins in determining this functional connectivity profile. Using functional magnetic resonance imaging (fMRI) on 820 subjects from the Human Connectome Project, and magnetoencephalographic (MEG) recordings from a subset, the heritability of connectivity among 39 cortical regions was estimated. On average over all connections, genes account for about 15% of the observed variance in fMRI connectivity (and about 10% in alpha-band and 20% in beta-band oscillatory power synchronisation), which substantially exceeds the contribution from the environment shared between twins. Therefore, insofar as twins share a common upbringing, it appears that genes, rather than the developmental environment, have the dominant role in determining the coupling of neuronal activity.

Highlights

Intrinsic human brain activity enables inference about the pathways and processes of information transfer in the brain
Using resting-state functional magnetic resonance imaging (fMRI) and MEG recordings released as part of the Human Connectome Project (HCP), we have constructed the functional network structure that expresses the coupling in MEG oscillatory power within three frequency bands, and fMRI partial correlation networks, among 39 regions of interest (ROIs)
We have shown that the entire functional network structure is found to be more similar for two subjects the more closely they are related, and we estimate that the average heritability of individual connection strengths is about 15–18% for BOLD correlations, and between 13% and 26% (1% and 25%) for correlations in beta-band power fluctuations

Summary

Introduction

Intrinsic human brain activity enables inference about the pathways and processes of information transfer in the brain. Studying intrinsic activity, when the brain is in a resting state, has given insights into many aspects of healthy and diseased brain function. Resting-state function is characterised by spatially separated regions organised into networks of strongly correlated activity (Beckmann et al, 2005; Smith et al, 2012). These networks represent both local connectivity and longer-range communication. Many neurological diseases, including schizophrenia and Alzheimer’s disease (Sheline and Raichle, 2013; Greicius, 2008), have been associated with major alterations to the strength and organisation of Colclough et al eLife 2017;6:e20178.

Methods

Results

Conclusion