Cortical thickness or grey matter volume? The importance of selecting the phenotype for imaging genetics studies

Abstract

Choosing the appropriate neuroimaging phenotype is critical to successfully identify genes that influence brain structure or function. While neuroimaging methods provide numerous potential phenotypes, their role for imaging genetics studies is unclear. Here we examine the relationship between brain volume, grey matter volume, cortical thickness and surface area, from a genetic standpoint. Four hundred and eighty-six individuals from randomly ascertained extended pedigrees with high-quality T1-weighted neuroanatomic MRI images participated in the study. Surface-based and voxel-based representations of brain structure were derived, using automated methods, and these measurements were analysed using a variance-components method to identify the heritability of these traits and their genetic correlations. All neuroanatomic traits were significantly influenced by genetic factors. Cortical thickness and surface area measurements were found to be genetically and phenotypically independent. While both thickness and area influenced volume measurements of cortical grey matter, volume was more closely related to surface area than cortical thickness. This trend was observed for both the volume-based and surface-based techniques. The results suggest that surface area and cortical thickness measurements should be considered separately and preferred over gray matter volumes for imaging genetic studies.

Introduction

Despite evidence that most neurological and psychiatric illnesses are genetically mediated, the identification of genes that predispose brain-related disorders has been difficult. One strategy for unraveling the genetic determinants of these complex illnesses is the use of endophenotypes, also called intermediate phenotypes (Gottesman and Shields, 1967, Gottesman and Gould, 2003), which are quantitative traits that are genetically correlated with illness. Neuroimaging methods provide an array of potential endophenotypes for these disorders (Glahn et al., 2007). Although there is mounting evidence that quantitative indices derived from brain structure and function are heritable (Baaré et al., 2001, Thompson et al., 2001, Geschwind et al., 2002, Peper et al., 2007, Glahn et al., 2007), and associated with neuropsychiatric disorders (McDonald et al., 2006, Honea et al., 2008, Goldman et al., 2008Goldman et al., 2009, van der Schot et al., 2009), it is unclear which genes influence these traits or the biological mechanisms that govern these measurements. The search for genes that influence brain-related traits could be improved by choosing imaging phenotypes that are closest to a single gene action (Wojczynski and Tiwari, 2008). However, given the complexity and multi-dimensional nature of imaging data, it is critical to determine the appropriate measurements to be employed. In this report, we examine issues surrounding the use of imaging-derived grey matter measurements in imaging genetics studies.

Although the human brain is gyrencephalic, with no absolute linear relationship between brain volume and surface area (Hofman, 1985, Armstrong et al., 1995), simple geometric laws still hold within the cortical mantle, where the grey matter volume is defined as the amount of grey matter that lies between the grey-white interface and the pia mater (Fig. 1, left). Previous findings suggest that cortical surface area and cortical thickness are independent, both globally and regionally, that grey matter volume is a function of these two indices and each of these three measurements are heritable (Winkler et al., 2009, Panizzon et al., 2009). Furthermore, Panizzon et al. (2009) reported that surface area and cortical thickness are genetically uncorrelated. Thus, volume measurements, which combine aspects of both traits, are likely influenced by some combination of these genetic factors. Since different definitions of a phenotype potentially lead to different genetic findings (Rao, 2008), it would seem that surface or thickness measures would have advantages over volume measures for gene discovery.

Cortical anatomy, which is structured as a corrugated two-dimensional sheet of tissue, can be well represented by surface models, which facilitate the analysis of relationships between cortical regions and provide superior visualisation (van Essen et al., 1998). Intersubject and even interspecies registration can be accomplished using surface-based representations (van Essen et al., 2001), allowing matching of homologies without relying directly on spatial smoothing as in volume-based methods (Ashburner and Friston, 2000). Computational advances in surface reconstruction (Mangin et al., 1995, Dale et al., 1999, Fischl et al., 1999a) and the availability of software packages also facilitates its use. Despite these advantages, quantitative methods based on purely volumetric representations of the brain are still common. Amongst these, voxel-based morphometry (VBM) (Wright et al., 1995, Bullmore et al., 1999, Ashburner and Friston, 2000, Good et al., 2001) is a relatively fast, straightforward method, that quantifies the amount of grey matter existing in a voxel and permits a comparison across subjects (Fig. 1, right). VBM requires that voxels are classified according to different tissue types, usually grey matter, white matter and cerebrospinal fluid (GM/WM/CSF). To allow comparison across subjects, images are non-linearly aligned to a standard brain, where a common coordinate system can be defined, and volumes are corrected for local shrinkages and expansions using the Jacobian determinants of the warps at each voxel (Good et al., 2001). In this way, VBM allows for the quantification of the grey matter volumes, globally and regionally, using either the voxels directly or regions of interest.

In this study, we compare measurements obtained with VBM methods, with measurements from surface-based representations of the brain in genetically informative samples. We focus our attention to determining the genetic control over (1) the volume of the grey matter computed using surface-based and volume-based representations of the brain, (2) the cortical surface area and (3) the cortical thickness.