Skip to main content
NCBI home page
Cerebral Cortex (New York, NY) logoLink to Cerebral Cortex (New York, NY)
. 2020 Jul 17;30(12):6191–6205. doi: 10.1093/cercor/bhaa174

Quantifying Genetic and Environmental Influence on Gray Matter Microstructure Using Diffusion MRI

Madhura Baxi 1,2,✉, Maria A Di Biase 2,#, Amanda E Lyall 2,3,#, Suheyla Cetin-Karayumak 2, Johanna Seitz 2, Lipeng Ning 2, Nikos Makris 2,3, Douglas Rosene 4, Marek Kubicki 2,3, Yogesh Rathi 2,3,5
PMCID: PMC7732156  PMID: 32676671

Abstract

Early neuroimaging work in twin studies focused on studying genetic and environmental influence on gray matter macrostructure. However, it is also important to understand how gray matter microstructure is influenced by genes and environment to facilitate future investigations of their influence in mental disorders. Advanced diffusion MRI (dMRI) measures allow more accurate assessment of gray matter microstructure compared with conventional diffusion tensor measures. To understand genetic and environmental influence on gray matter, we used diffusion and structural MRI data from a large twin and sibling study (N = 840) and computed advanced dMRI measures including return to origin probability (RTOP), which is heavily weighted toward intracellular and intra-axonal restricted spaces, and mean squared displacement (MSD), more heavily weighted to diffusion in extracellular space and large cell bodies in gray matter. We show that while macrostructural features like brain volume are mainly genetically influenced, RTOP and MSD can together tap into both genetic and environmental influence on microstructure.

Keywords: imaging genetics, MSD, non-Gaussian model, RTOP, twin study

Introduction

Twin studies provide a powerful means to estimate the degree of genetic and environmental influence on a certain trait (e.g., height, weight, behavior, cognition, and brain structure). These estimates are derived by comparing the genetic and environmental overlap within and between monozygotic twins (MZ), dizygotic twins (DZ), and nontwin siblings. For example, greater phenotypic overlap in MZ twins relative to DZ twins and siblings indicates greater genetic influence on that given phenotype. In contrast, little or no difference in phenotypic overlap in MZ compared with DZ twins and siblings implies that shared environmental factors primarily influence phenotypic outcomes. Additionally, the impact of unique environmental experiences (e.g., individual lifestyle, diet, exercise, stress, relational experience, and epigenetics) can be inferred from differences between MZ twins. In this way, twin designs enable estimates of genetic, shared environmental, and unique environmental influences on phenotypes.

Several previous studies have employed twin designs to study the genetic and environmental influence on brain structure. These studies have primarily focused on structural magnetic resonance imaging (sMRI)-based macrostructural brain imaging phenotypes, such as cortical thickness, surface area, and gray/white matter volume, which are predominantly (up to 90%) influenced by genetic factors (Peper et al. 2007; Schmitt et al. 2007; Kremen et al. 2012; Jansen et al. 2015; Lukies et al. 2017). Conversely, less is known about the influence of genetic and environmental factors on the underlying microstructural properties of gray matter (GM), such as the architectural composition of cell bodies and cellular processes. Elucidating genetic and environmental impacts on brain microstructure is important as several psychiatric and neurological disorders such as schizophrenia, depression, Parkinson’s, and Alzheimer’s exhibit an array of intra- and extracellular GM pathologies (Dickson et al. 2012; Rajkowska and Stockmeier 2013; Athanas et al. 2015; Weston et al. 2015; Nazeri et al. 2017). Furthermore, all of these disorders are affected by both genes (Tanner et al. 1999; Bekris et al. 2010; Gejman et al. 2010; Franke et al. 2016; Shadrina et al. 2018) and environmental factors (Dean and Murray 2005; Landrigan et al. 2005; Demjaha et al. 2012). Thus, it is important to understand the proportion of genetic and environmental influence on GM microstructure as estimated using diffusion-magnetic resonance imaging (dMRI)-derived measures.

Microstructural properties can be measured from dMRI, which is sensitive to changes in the underlying organization of both white matter (Hagmann et al. 2006; Le Bihan and Iima 2015) and GM tissue (Rathi et al. 2014; Weston et al. 2015; Seitz et al. 2018). Since dMRI has been predominantly used to study white matter, only a few studies have addressed the influence of genetic and environmental factors on microstructure-sensitive measures in GM. Of those few looking at GM microstructure, most employed conventional dMRI measures, such as the mean diffusivity (MD), computed from single tensor models (i.e., diffusion tensor imaging, DTI). For example, Elman et al. (2017) used the Vietnam Era Twin Study of Aging twin sample of 420 subjects (age range: 56–66 years) and conventional DTI, which revealed a high influence of genes (65%) and low to moderate influence of environment (35%) on the MD of whole brain cortical GM. This would suggest that both genes and environment can impact GM microstructure. It is important to note, however, that these measures are too simple to represent the complex GM microstructure. DTI model assumes that water molecular diffusion follows a monoexponential Gaussian distribution, implying that water diffusion is free from restrictions and hindrances (Basser et al. 1994; Jones et al. 2010; O’Donnell et al. 2011). It is known, however, that the water in biological structures displays much more complex behavior due to the presence of restrictions and barriers from myelin and cell membranes and thus, always exhibits non-Gaussian behavior (Kärger et al. 1988; Jensen et al. 2010).

Several advanced dMRI models have been proposed to overcome the limitations in DTI (Mulkern et al. 1999; Özarslan et al. 2013; Rathi et al. 2013) in better defining underlying cytoarchitecture. One such method that allows for analytical calculation of the three-dimensional probability distribution of water diffusion in tissue is given by a biexponential model (Mulkern et al. 1999; Rathi et al. 2013). This model requires data from advanced, multishell dMRI acquisition and is more sensitive to restricted water diffusion within brain tissue (Özarslan et al. 2013; Ning et al. 2015). Two scalar measures that can be computed using this advanced dMRI model are return to origin probability (RTOP) and mean squared displacement (MSD). RTOP and MSD have been successfully used to investigate brain microstructure in in-vivo healthy human studies and in ex-vivo experiments on a marmoset brain (Özarslan et al. 2013; Avram et al. 2016; Fick et al. 2016). Avram et al. (2016) reported that even though single–shell-based DTI is the most commonly used method to noninvasively characterize architectural features of brain tissue, the assumption of Gaussian distribution of water diffusion weakens its ability to describe intricate tissue microstructure. Indeed, multishell advanced dMRI measures, for example, RTOP and MSD provide a more comprehensive tissue characterization as they provide more specific information about cellularity, size of cell bodies, and its processes and presence of restricting barriers (e.g., myelin; Avram et al. 2016). Specifically, RTOP has been shown to reflect cellularity and restrictions better than MD and is thus more sensitive to complex diffusion processes, for example, effects of restricting barriers (Avram et al. 2016). On the other hand, MSD is considered to be mainly sensitive to fast diffusion, for example, in extracellular spaces and large cell bodies (Boscolo Galazzo et al. 2018). Very few studies have investigated advanced dMRI measures in clinical applications, albeit showing promising potential to detect brain alterations in attention-deficit/hyperactivity disorder (Wu et al. 2019), ischemic stroke (Boscolo Galazzo et al. 2018), and Alzheimer’s disease (Fick et al. 2016). For example, Boscolo Galazzo et al. (2018) investigated the sensitivity of advanced dMRI measures to stroke-induced microstructural pathologies, and reported better sensitivity and specificity of MSD over MD in identifying and characterizing different portions of the ischemic lesion (Boscolo Galazzo et al. 2018). Further, a recent study (Pines et al. 2020) conducted in 120 young adults (ages 12–30) showed higher sensitivity of the multishell diffusion imaging-derived measures to age effects compared with the DTI measures. Additionally, the study also reported that RTOP is less vulnerable to in-scanner head motion which is known to be a major confound for neurodevelopmental studies, since motion related are exacerbated in younger cohorts with psychiatric disorders. As such, based on previous studies, it is apparent that multishell imaging-derived advanced dMRI measures of RTOP and MSD have great promise and important advantages over DTI measures, however, the degree of genetic and environmental influence on these properties has not been explored, which is the focus of this paper.

Therefore, in this study, we leverage publicly available multishell dMRI data from the Human Connectome Project (HCP) to examine genetic, family-wide shared and unique environmental influences on micro- and macrostructural properties of GM in healthy adults. Specifically, we examine advanced dMRI measures of RTOP and MSD, as well as macrostructural sMRI measures of cortical thickness and GM volume.

Methods

Demographics

This study included 840 healthy adult subjects (471 females, 369 males, age range = 22–37 years, age mean = 28.8 years, standard deviation (SD) = 3.66) drawn from the HCP (Van Essen et al. 2013), which underwent ethics approval. Participants included 170 MZ twins (50 males/120 females, age mean = 29.7 years, age SD = 3.36), 162 DZ twins (72 males/90 females, age mean = 29.1 years, age SD = 3.15), 413 siblings (203 males/210 females, age mean = 28.4 years, age SD = 3.82), and 95 unrelated individuals with no twins or siblings in the study (44 males/51 females, age mean = 28.3 years, SD = 3.96).

Acquisition Information and Preprocessing

Detailed descriptions of the acquisition parameters and preprocessing pipelines are provided in the HCP documentation (Glasser et al. 2013; Sotiropoulos et al. 2013; Uğurbil et al. 2013; Van Essen et al. 2013). In brief, high-resolution T1 and multiple b-value (multishell) diffusion MRI scans (Fig. 1) were acquired on a 3 T Siemens connectome scanner with the following parameters: T1 MPRAGE: TR/TE/TI = 2400/2.14/1000 ms, flip angle = 8°, FOV = 224 × 224 mm2, voxel size = 0.7 mm3 isotropic, bandwidth = 210 Hz/pixel. Multishell dMRI: Spin-echo EPI, TR/TE = 5520/89.5 ms, flip angle = 78°, FOV = 210 × 180 mm2, voxel size = 1.25 mm3 isotropic, and 3 b-values: 1000, 2000, and 3000 s/mm2. A full dMRI session included six runs representing three different gradient tables. Each gradient table included 90 diffusion-weighted directions plus 6 b = 0 acquisitions interspersed throughout each run.

Figure 1.

Figure 1

Open in a new tab

Pipeline for T1 and diffusion MRI analysis. Steps of this analysis included, masking and brain extraction of preprocessed T1 and multishell diffusion MRI (DWI) image; Freesurfer anatomical parcellation of T1 images; fitting a biexponential model to DWI images resulting in advanced dMRI scalar measures of MSD and RTOP maps; computing regional MSD and RTOP measures using the anatomical ROIs obtained from Freesurfer parcellations registered to MSD and RTOP maps; extracting macrostructural measures of average regional cortical thickness and gray matter volume using the Freesurfer parcellation; and estimating the genetic and environmental influence on these average regional measures using the SOLAR package.

DMRI preprocessing as detailed in the HCP documentation (Glasser et al. 2013) included the following steps: b0 intensity normalization, eddy current correction, motion correction, EPI distortion correction, gradient nonlinearity correction, registration to T1 in MNI space, and brain extraction. T1 scans were preprocessed using the following steps: gradient distortion correction, brain extraction, readout distortion correction, bias field correction, and nonlinear registration to MNI space (Glasser et al. 2013).

Macrostructural GM sMRI Measures

Preprocessed T1 scans underwent brain parcellation into GM cortical and subcortical regions of interest (ROIs; Fig. 1) using the Desikan–Killiany cortical atlas (Desikan et al. 2006) and automatic subcortical segmentation (Fischl et al. 2002) using FreeSurfer software version 5.2 (http://surfer.nmr.mgh.harvard.edu/). T1 images were also used to compute regional cortical volume and thickness estimates in FreeSurfer (Fischl and Dale 2000; Pengas et al. 2009; Keller et al. 2012). Additionally, the ratio of regional GM volume to whole brain volume (GM volume ratio) was computed for all ROIs (Table 1) to remove the effect of intersubject head size variance on regional/global GM volume.

Table 1.

Measures and gray matter ROIs used in this study

Measures ROIs
Macrostructural measures: 1. Whole-brain cortical gray matter
1. Cortical thickness 2. Hemispherical whole-brain cortical gray matter
2. GM volume to whole-brain volume ratio (GM volume ratio) 3. Regional cortical gray matter: Left and right frontal, parietal, temporal, occipital lobes, and hippocampus
Microstructure-sensitive diffusion measures: 4. Subcortical gray matter: Left and right striatum, amygdala, thalamus, globus pallidum, and ventral DC
1. Return to origin probability
2. Mean squared displacement
Open in a new tab

Note: GM, gray matter; RTOP, return to origin probability; MSD, mean squared displacement; ROI, region of interest.

Using the FreeSurfer parcellations, average values of abovementioned macrostructural measures (cortical thickness and GM volume ratio) and microstructure-sensitive diffusion measures (RTOP and MSD) described in the next subsection were computed in the following ROIs (Table 1): (1) whole brain cortical GM; (ii) whole brain left and right hemispherical GM; (iii) left and right frontal, parietal, temporal, occipital, hippocampal GM; and (iv) left and right subcortical regions of striatum, amygdala, thalamus, globus pallidus, and ventral diencephalon (DC).

Microstructure-Sensitive GM Advanced Diffusion MRI Measures

A biexponential model was fit to the multishell diffusion data in GM regions, which allowed modeling of non-Gaussian water diffusion in brain tissue (Özarslan et al. 2013; Rathi et al. 2013). This model provides more sensitive measures for GM microstructure than conventional DTI measures, such as FA and MD (Assaf et al. 2002). More specifically, for GM areas, the model consisted of an isotropic compartment to model cerebrospinal fluid (CSF) contamination (Diso = 3 μm2/ms) and a restricted compartment to model the non-Gaussian diffusion in the tissue in each voxel (Rathi et al. 2013). A biexponential function was used to model the water diffusion in the restricted tissue compartment (in the GM, we assumed a single restricted compartment to capture non-Gaussian diffusion due restrictions in small cell bodies or intra/extracellular spaces). This model allows estimating the three-dimensional probability distribution of water molecule displacement at each voxel (often referred to as the diffusion propagator) (Mitra and Halperin 1995; Özarslan et al. 2013; Rathi et al. 2013). Advanced dMRI scalar measures of RTOP and MSD were then computed for each voxel after removing CSF contamination (i.e., removing the isotropic compartment from each measure).

MSD and RTOP maps (Fig. 1) were computed for each subject as follows:

graphic file with name M1.gif

where r is the displacement vector and Inline graphic (where E(q) is the diffusion MRI signal along a q-space point q), is the average diffusion propagator (Ning et al. 2015). MSD primarily captures the displacement of fast-moving water molecules. An example axial slice of an MSD map is shown in Figure 1. In GM, MSD can be affected by hindrances in the extracellular space, such as variations in cell size or density.

RTOP is the probability that a water molecule returns to its starting position in a given experimental diffusion time. More restriction leads to higher RTOP. RTOP primarily captures information about slowly diffusing water, such as in the restricted intracellular and intra-axonal spaces.

Test–Retest Reliability of RTOP and MSD in GM

As part of the HCP study, a subset of MZ twins (n = 45, age range: 22–35 years) were scanned again with the test–retest time interval ranging from 2 to 11 months, using the same scanning parameters as the first visit scans (Van Essen et al. 2013). The same processing pipeline was used for the retest data as well, with the biexponential model used to fit the multishell dMRI data. RTOP and MSD measures were calculated for the whole brain cortical GM. We then conducted test–retest reliability analysis using two commonly used statistical methods: (1) Percent test–retest within subject variability (PTRT) (Baumgartner et al. 2018). (2) Coefficient of variation (CoV) (Brian Everitt 1998; Brown 1998; Lovie 2005; Baumgartner et al. 2018). These statistical methods provide us information about how close the measurements (RTOP and MSD) are when obtained repeatedly on the same subject under identical conditions (same scanning protocol). Closer the test–retest measurements, lower the PTRT and CoV and hence higher the test–retest reliability of the measures.

Modeling the Genetic and Environmental Impact on GM Structure

All statistical analyses were conducted using R software version 3.3.3 (R Core Team 2012). Subjects for which either of the whole brain GM measures (RTOP, MSD) was >3 SD from their population mean were considered outliers (n = 19, 0.03% of all subjects) and were removed from the analysis.

Pairwise Correlation: Descriptive Statistics

To assess phenotypic similarity for each imaging measure within different subject groups, Pearson pairwise correlations were computed for whole-brain-averaged cortical GM measures (Fig. 1) separately within: (i) MZ twin pairs, (ii) DZ twin pairs, (iii) sibling pairs, and (iv) unrelated subjects (Table 2). We conducted these Pearson correlations using bootstrapping which involved computing Pearson correlation for 1000 iterations of resampling and swapping 20 subjects between groups 1 and 2 in each iteration. Upon bootstrapping, mean Pearson correlation and its SD were computed for each group (Table 2). These correlations were used to estimate the combined (genetic + environmental) similarity of a phenotypic trait within a pair in the next step.

Table 2.

Pairwise correlation between twins, nontwin siblings, and unrelated pairs: whole brain cortical measures

Type Cortical thickness GM volume ratio RTOP MSD
MZ twins 0.78* ± 0.01 0.94* ± 0.004 0.42* ± 0.01 0.23 ± 0.01
DZ twins 0.53* ± 0.01 0.72* ± 0.01 0.30* ± 0.01 0.03 ± 0.01
Nontwin siblings 0.33* ± 0.01 0.46* ± 0.01 0.27* ± 0.01 0.11 ± 0.01
Unrelated 0.07 ± 0.02 0.02 ± 0.02 0.01 ± 0.02 −0.14 ± 0.02
Open in a new tab

* indicates statistical significance (P < 0.05).

SOLAR ACE Model Analysis

To separately quantify genetic and environmental components of a phenotypic trait, the Sequential Oligogenic Linkage Analysis Routines (SOLAR) package was used to decompose the variance of each phenotypic brain imaging trait into three components: genetic (A), common environmental (C) and unique environmental (E) influence on the phenotype (ACE model; Tenesa et al. 2013) (Almasy et al. 1998, www.solar-eclipse-genetics.org), as described below:

ACE model: A (additive genetics) C (common environment) and E (unique environment)

graphic file with name M3.gif

where y is the phenotype of one of the dMRI measures: regional RTOP, MSD, cortical thickness, or GM volume. X is the covariate (e.g., age, sex, age × sex in this work), N is the multivariate Gaussian, Φ is a kinship matrix between subjects, σa2 is the variance attributable to genetics, γ is a matrix of the common environment between subjects, σc2 is the common environmental variance, σe2 is the unique environmental variance, Φ codes for the expected fraction of genome shared between subjects (i.e., identical, MZ twins as 1, DZ twins, and sibling–sibling as 0.5, parent–child as 0.25, grandparent–grandchild as 0.125, etc.). Note that γ represents an assumed common environment between relationship pairs. Specifically, γ is set to 1 for the same household, and otherwise to 0.

The genetic component (A), common environmental component (C), and unique environmental component (E) were computed after covarying out the effects of age, sex, and the interaction between age and sex. GM volume was not included as a covariate in our heritability analysis of RTOP, since no significant correlations were found between GM volume and GM RTOP at whole brain level. Each estimated component can range from 0 to 1 and all three components will sum to 1. It is important to note that if a genetic or environmental component of a phenotype is, for example, 0.7 it does not mean that a trait is 70% caused by genetic or environmental factors; it means that 70% of the variability in the trait in a population is due to genetic or environmental influence.

The ACE model was fit to whole-brain and regional GM brain measures (MSD, RTOP, cortical thickness, and the GM volume ratio), these ROIs are listed in Table 1. The ACE model estimated the genetic and environmental components of a phenotypic variance along with the standard errors and P-values for these estimates. False discovery rate was applied to correct for multiple comparisons across all the ROIs and brain measures (listed in Table 1). The SOLAR package also was used to estimate coefficients for each covariate (e.g., age) of the phenotype examined, along with its standard errors and P-values. Additionally, we used the data homogenization approach (implemented in solar-eclipse; www.solar-eclipse-genetics.org proposed by Kochunov et al. 2019b) to improve the convergence of heritability estimates across different methods. Homogenization involved enforcing normality on the trait data using inverse Gaussian transformation and accounting for the appropriate covariates such as age and sex. This step was performed before fitting the ACE model to the imaging phenotypes.

Results

Test–Retest Reliability Estimates

Based on the results of our test–retest reliability analysis, RTOP in GM showed low PTRT of 2.88% and CoV of 2.04%. Similarly for MSD in GM, our results showed low PTRT and CoV of 1.99% and 1.41%, respectively.

Descriptive Statistics

Phenotypic similarity across groups: Pearson pairwise correlations were computed in each group for whole-brain cortical thickness, GM volume, RTOP, and MSD to assess phenotypic similarity. Cortical thickness and GM volume were strongly correlated in MZ twin pairs (cortical thickness: r = 0.78; P = 3.3*10−18; GM volume: r = 0.94, P = 7.1*10−41) and less concordant in DZ twins (cortical thickness: r = 0.53 P = 7.9 × 10−7; GM volume: r = 0.72, P = 4.6 × 10−14 and siblings (cortical thickness: r = 0.33, P = 0.0013; GM volume: r = 0.46, P = 3 × 10−6, indicating that genetic factors strongly contribute to macrostructural phenotypes at whole-brain level (Table 2). RTOP was moderately correlated in MZ twin pairs (r = 0.42, P = 3.1 × 10−4), DZ pairs and siblings (r = 0.30, P = 0.01; r = 0.27, P = 0.02, respectively), indicating contributions of both genetic and environmental factors (Table 2). In contrast, MSD was weakly correlated in MZ, DZ pairs, and siblings (r = 0.23, P = 0.13; r = 0.03, P = 0.82; r = 0.11, P = 0.45, respectively), which can be interpreted as MSD being predominantly influenced by unique environmental factors (Table 2).

SOLAR Model Genetic and Environmental Influence Estimates

ACE Modeling of GM Macrostructure

SOLAR estimates for the macrostructural measures of cortical thickness and GM volume ratio at a whole-brain level, showed high influence of genetic factors (0.81 and 0.76, respectively; P < 0.05; Fig. 2A), with little to no unique and common environmental effects (P > 0.05; Fig. 2A). Similar results were obtained for each hemisphere (Fig. 2B), and across the majority of cortical (Fig. 3A) and subcortical (Fig. 4A) brain regions, demonstrating a predominant genetic influence on macrostructural GM measures (range = 0.55–0.89, P < 0.05; Figs 2B, 3A, and 4A) with little to no effect of unique or shared environment.

Figure 2.

Figure 2

Open in a new tab

Genetic and environmental influence on imaging measures in whole brain and hemispherical cortical gray matter. (A) Whole brain cortical GM: Table displays estimates of genetic (A), unique environmental (E), and common environmental influence (C) on whole brain cortical GM average macrostructural and microstructure-sensitive dMRI measures, along with the corresponding standard error and P-values. It also presents regression coefficients and P-values for each covariate (age, sex, age × sex interaction term) included in the SOLAR model. (B) Hemispherical cortical GM: Wheel plot presents SOLAR model estimates of genetic (inner rim), unique environmental (middle rim), and common environmental (outer rim) influence on macrostructural (left half) and microstructure-sensitive advanced dMRI measures (right half). Estimated influence of genetic and environmental factors on the imaging measures in left and right cortical GM has been color-coded as indicated in the color-bar below the wheel plot (ranging from 0: light yellow to 1: dark red).

Figure 3.

Figure 3

Open in a new tab

Genetic and environmental influence on imaging measures in cortical GM regions. Wheel plots show SOLAR model estimates of genetic (inner rim), unique environmental (middle rim), and common environmental (outer rim) influence on (A) macrostructural measures of GM volume ratio (left half) and cortical thickness (right half), and (B) microstructure-sensitive advanced dMRI measures of RTOP (left half) and MSD (right half). Estimated influence of genetic and environmental factors on the imaging measures in left and right cortical GM regions has been color-coded as indicated in the color-bar below the wheel plot (ranging from 0: light yellow to 1: dark red).

Figure 4.

Figure 4

Open in a new tab

Genetic and environmental influence on imaging measures in subcortical GM regions. Wheel plots show SOLAR model estimates of genetic (inner rim), unique environmental (middle rim), and common environmental (outer rim) influence on (A) macrostructural measures of GM volume, and (B) microstructure-sensitive advanced dMRI measures of RTOP (left half) and MSD (right half). Estimated influence of genetic and environmental factors on the imaging measures in left and right subcortical GM regions has been color-coded as indicated in the color-bar below the wheel plot (ranging from 0: light yellow to 1: dark red).

ACE Modeling of GM Microstructure

RTOP for whole brain cortical GM and for both hemispheres showed moderate genetic influence (range = 0.45–0.54; P < 0.05; Fig. 2A,B), moderate unique environmental influence (range = 0.46–0.55; Fig. 2A,B) and little to no common environmental effects (P > 0.05). These results were mirrored in the majority of cortical and subcortical regions with moderate influence of genetic factors (range = 0.25–0.52; P < 0.05 Figs 3B and 4B), moderate to high unique environmental factors (range = 0.44–0.82; Figs 3B and 4B), and little to no common environmental influence (range = 0–0.18; P > 0.05; Figs 3B and 4B). Relative to the other cortical regions, the hippocampus showed lower genetic influence (0.07–0.33; P > 0.05; Fig. 3B).

MSD for whole brain cortical ROI in both hemispheres displayed little to no genetic influence (0–0.12, P > 0.05), high unique environmental influence (range = 0.77–0.88) (Fig. 2A,B) and very low common environmental influence (range = 0.08–0.23; P > 0.05). Similar results were observed for a majority of cortical lobes, that is, low genetic influence (range = 0–0.3; P > 0.05), high unique environmental effects (range = 0.57–0.83; Fig. 3B) and low common environmental effects (range = 0–0.22; P > 0.05). However, MSD in the hippocampus showed slightly different trend, exhibiting a moderate genetic influence (range = 0.33–0.43; P < 0.05; Fig. 3B). Similarly, majority of subcortical regions also displayed slightly different results for MSD compared with the cortical regions, with moderate genetic influence (range = 0.3–0.48; P < 0.05), moderate to high unique environmental effects (range = 0.5–0.7), and low common environmental influence (range = 0–0.15; P > 0.05; Fig. 4B).

Discussion

In this study, we investigated the influence of genes and environment on advanced dMRI measures of RTOP and MSD by leveraging the high-resolution multishell dMRI data obtained from a large twin and sibling population (840 subjects; age range: 22–37 years), as part of the HCP. The two main findings of this study are the following: First, while sMRI measures were mainly affected by genetic factors, microstructure-sensitive advanced dMRI measures in GM showed influence of both genetic and unique environmental factors. Second, we demonstrated for the first time that the advanced dMRI measures of RTOP and MSD have the ability to noninvasively tap into regionally specific patterns of genetic and environmental influence in neocortical and subcortical GM.

Additionally, based on our test–retest reliability analysis, it is clear that the advanced dMRI measures of RTOP and MSD in GM are reliable and reproducible, since they both showed low PTRT and CoV of <3% (Shahim et al. 2017; Baumgartner et al. 2018). Thus, the contribution of variance to unique environment in the ACE model is not driven by random noise but is reproducible. Further, RTOP and MSD are known to be more sensitive to underlying microstructure compared with DTI measures and hence provide better representation of the underlying architecture of GM.

Macrostructure Is Mainly Genetically Driven While Microstructure Is Shaped by Both Genes and Environment

Our results, consistent with previous findings (Thompson et al. 2001; Joshi et al. 2011; Jansen et al. 2015), demonstrate that sMRI measures of cortical thickness and GM volume were predominantly determined by genetic factors. On the other hand, advanced dMRI measures were overall influenced by a combination of both genetic and environmental factors. Since sMRI and dMRI tap into different biophysical information, it could explain why the influence of genes and environment shows different patterns on the measures derived from these two modalities. Our findings suggest that while gross anatomy of GM might be driven mainly by genes, the underlying microstructure might be influenced by both genes and environment. This is consistent with reviews, supporting the view that genetic and environmental factors both contribute to the pattern and degree of gyrification and changes in related microstructural properties, whereas gross anatomical feature of brain size is more strongly driven by genetic factors (Bartley et al. 1997; Zilles et al. 2013).

The findings from the present study confirm a generally accepted concept that the combination of genetic and environmental factors shapes the final pattern of GM microstructure. Rakic 1988 discussed this concept in the context of cortical expansion during neurodevelopment, named “proto-map.” The prefix “proto” was added to emphasize that the formation of cytoarchitectonic areas is malleable, affected also by environmental factors, even though the initial map is genetically determined. Similarly, some other studies have discussed this concept in the context of gyrification, supporting the roles of the genes in driving gyrification and related microstructural properties, for example, neuronal migration (Bartley et al. 1997; Rogers et al. 2010; Zilles et al. 2013; Kroenke and Bayly 2018), while also highlighting the impact of forces during development (environmental factors) on folding of cerebral cortex (e.g., Duque et al. 2016; Kroenke and Bayly 2018). Several lines of evidence have shown that the impact of environmental insults, for example, intrauterine growth restriction (Dubois et al. 2008), premature birth (Zhang et al. 2015) and fetal alcohol exposure (Hendrickson et al. 2017) can lead to abnormal folding patterns (Rakic and Swaab 1988; Kroenke and Bayly 2018). Therefore, previous studies suggest that brain developmental processes including gyrification, neuronal migration, patterned neurogenesis, and cell differentiation might be regulated by an intricate relationship between genetics and environmental factors (Llinares-Benadero and Borrell 2019). Our finding of GM microstructure being influenced by both genes and environment is thus consistent with the abovementioned literature on brain development.

On the other hand, as shown by our and many previous studies, macrostructural measures of cortical thickness and GM volume are predominantly influenced by genetic factors and are hence limited in tapping into environmental influences. Moreover, as stated by a recent review paper, our current knowledge lacks understanding of the impact of environmental factors on GM microstructure, as the focus has mostly been on understanding the genetically driven mechanisms driving microstructural changes in the brain (Quezada et al. 2018). The present study thus adds much needed additional evidence to the concept of GM microstructure being shaped by not only genes, but also an individual’s unique environment.

Advanced dMRI Measures Highlight the Differential Role of Both Unique Environment and Genes in Shaping Microstructure of Different GM Regions

Our results also suggest that the advanced dMRI measures are not only sensitive to tissue microstructure, but that the two different measures of RTOP and MSD also represent distinct biophysical properties of the underlying tissue microstructure. RTOP primarily captures restricted water diffusion such as in intracellular and intra-axonal spaces (e.g., water displacement <5 μm) and thus is a likely indicator of cellularity, size of cell bodies and its processes in GM. On the other hand, MSD is predominantly sensitive to diffusion of water molecules in larger spaces (e.g., >7 μm), such as the extracellular spaces or intracellular spaces within large cell bodies.

In this study, RTOP showed a moderate influence of both genetic and unique environmental factors on a majority of cortical and subcortical GM regions. On the contrary, MSD showed regionally specific effects: MSD in neocortical brain regions was almost entirely driven by unique environmental factors with the exception of hippocampus, which showed a moderate influence of both genes and unique environment. Our results thus unveiled important differences between hippocampus and rest of the cortex in the influence of genes and environment. It is well documented that the hippocampus is a complex structure known to be functionally, cytoarchitecturally, and anatomically quite different from the rest of the cortex (McClelland et al. 1995; Shankle et al. 1998; Grove and Tole 1999; Shaw et al. 2008; Norman 2010; Bay et al. 2018). Additionally, hippocampus is known to follow different developmental trajectory compared with neocortex (Shankle et al. 1998; Grove and Tole 1999; Shaw et al. 2008). Due to its evolutionarily older and simpler architecture compared with neocortex, hippocampus has been suggested to be more vulnerable to both genetic and environmental perturbances (Shankle et al. 1998). For example, hippocampus is one of the earliest regions to be affected during Alzheimer’s disease. This literature is thus consistent with our finding of distinct pattern of moderate influence of both genes and environment on MSD in hippocampus compared with neocortical regions which showed mainly environmental influence on MSD.

Further, in subcortical regions, MSD showed a moderate influence of both genetic and unique environmental factors similar to hippocampus but unlike in neocortical regions which showed mainly unique environmental influence. These differences between neocortical and subcortical regions in the pattern of genetic and environmental influence are consistent with literature showing cytoarchitectural differences between these regions (Amunts and Zilles 2015). Subcortical regions are evolutionary more primitive similar to hippocampus than neocortical regions and have very different cytoarchitecture and cortical layers (Strominger et al. 2012). Moreover, higher genetic conservation of cytoarchitecture in subcortical regions compared with more evolved neocortical regions intra (Amunts et al. 2004; Amunts et al. 2005; Amunts and Zilles 2015) and inter (Oldham et al. 2006) species have also been reported by previous studies. Evolutionary hierarchy known from previous literature is thus consistent with our finding of the pattern of higher genetic influence on MSD in subcortical regions compared with neocortical regions.

Previous studies have demonstrated an impact of unique environment on intracellular and extracellular properties of GM microstructure. For example, chronic stress can lead to the shrinkage of dendritic processes and suppressed synaptic input (McEwen et al. 2016). Furthermore, physical activity such as running can increase the neuronal spine density, synaptic plasticity, neurotrophin levels, and spatial memory function in mice (Voss et al. 2010). Additionally, extracellular matrix has been shown to undergo significant changes due to the influence of unique environmental factors such as training and exercise (Timmons et al. 2005; Bonnans et al. 2014; Guzzoni et al. 2018). These studies thus support our finding that unique environment plays a potentially substantial role in shaping the cellular composition and extracellular matrix. We also observed that MSD in GM overall showed more environmental influence compared with RTOP, which suggests that extracellular matrix properties might be affected more by unique individual experiences compared with intracellular properties. However, future postmortem studies are needed to directly compare genetic and environmental influence on extracellular versus intracellular properties. Nevertheless, this study for the first time provides the evidence that advanced dMRI measures of RTOP and MSD have the ability to capture unique environmental as well as genetic influence on intracellular and extracellular properties of GM microstructure.

Future Implications for Advanced dMRI Measures in Healthy and Clinical Conditions

Given the sensitivity of advanced dMRI measures of RTOP and MSD to the effects of individual experiences on GM microstructure, these measures could be used to monitor lifestyle impacts on health (e.g., stress or smoking) and can eventually improve person-specific treatments. Furthermore, previous studies conducted in healthy subjects have reported unique environmental influence on higher order cognitive functions associated with cortical regions such as attention, verbal/language processing, memory and visuo-spatial processing (Kramer et al. 2004; Tucker-Drob et al. 2013; Gustavson et al. 2018; Zhang and Xiaobo 2018). These studies highlight the crucial role of individual experiences in defining cortical GM microstructure and in turn, the associated behaviors or cognitive abilities.

Additionally, as we show in our study, macrostructural measures represent one end of the spectrum, as they are predominantly influenced by genes. The inclusion of these two new measures expands the present scope of imaging studies to not only study the impact of genes, but also the impact of the environment. Most psychiatric disorders as well as cognitive and behavioral changes during development and aging have been reported to be impacted by both genetic and environmental factors (Landrigan et al. 2005; McEwen and Getz 2013). Links between psychiatric symptoms as well as cognition and changes in GM microstructure have been well documented (Athanas et al. 2015; Reas et al. 2018). Collectively, all this evidence stresses the importance of studying the impact of both genes and environment on brain structure. As we discussed above, there is a very limited number of measurements available that allow for the assessment of both environmental and genetic impacts. Using RTOP and MSD, in conjunction with macrostructural measures, could, therefore, enable their application to a broader spectrum of behavioral, cognitive functions, and psychiatric symptoms that have been linked with both genes and environment.

Limitations

Findings from this study must be interpreted in the context of following potential limitations. It is important to note that MSD and RTOP are only indirect measures of underlying architecture in GM and, therefore, need biological validation. However, successful use of these measures in patients with ischemic stroke and Alzheimer’s animal model by previous studies (Boscolo Galazzo et al. 2018) provides clinical viability of these measures. Furthermore, estimates of genetic and environmental components using the ACE model are based on phenotypic variance across the population and not directly on genomic information. However, several studies have validated this model using genome-wide complex trait analysis and genome-wide association studies, which link imaging measures to specific single nucleotide polymorphisms (Koran et al. 2014; Guen et al. 2019; Kochunov et al. 2019a).

Conclusion

In summary, we conclude that individual-specific environment plays a key role in defining brain microstructure. While macrostructural brain features appear to be affected mainly by genetic factors, microstructural measures are more driven by the environment, which possibly contributes to the observed discordance in behavior and mental health between identical twins. Our study provides evidence of the capability of advanced dMRI measures including RTOP and MSD, in jointly capturing not only genetic but also unique environmental influences on brain microstructure. These measures, thus, have the potential for use in a wide range of studies that involve healthy and clinical population and aim to investigate/track changes in GM microstructure linked with both genes and environment.

Funding

We gratefully acknowledge the funding provided by following grants: RO1 AG042512 National Institute on Aging (PI: Dr Marek Kubicki, Dr Nikos Makris and Dr Doug Rosene); R01 MH111917 National Institute of Mental Health (PI: Dr Yogesh Rathi); R01 MH102377 National Institute of Mental Health (PI: Dr Marek Kubicki); R01 MH112748 National Institute of Mental Health (PI: Dr Marek Kubicki and Dr Nikos Makris); R01 MH119222 National Institute of Mental Health (PI: Dr. Yogesh Rathi); K01 MH115247 National Institute of Mental Health (PI: Dr Amanda Lyall); K24 MH110807 National Institute of Mental Health (PI: Dr Marek Kubicki).

Conflict of Interest: The authors declare that there is no conflict of interest.

A. Appendix

A.1. Biexponential model: CSF and restricted compartment (Rathi et al. 2013; Ning et al. 2015)

Diffusion signal S as a function of b-values (b) and diffusion gradient directions (u), was determined using:

graphic file with name M4.gif
graphic file with name M5.gif
graphic file with name M6.gif

Such that Inline graphic, Inline graphic, Inline graphic, Inline graphic and Inline graphic, Inline graphic, where Inline graphicis the weight fraction of the compartment; Inline graphic is the weight of gaussian in each biexponential; m, p, and q are orthogonal eigenvectors; and d0 is the diffusion coefficient of isotropic CSF compartment fixed to Inline graphic.

A.2. Diffusion pdf (propagator) of the biexponential model (Rathi et al. 2013; Ning et al. 2015)

Diffusion propagator for biexponential model has the following analytical form:

graphic file with name M16.gif

where Inline graphic and Inline graphic

A.3. RTOP and MSD: Biexponential model equations (Rathi et al. 2013; Ning et al. 2015)

Advanced dMRI measures of RTOP and MSD were derived using the P(r) (eq. A.2). Following are the analytical formulae for the measures:

graphic file with name M19.gif
graphic file with name M20.gif

In this work, we removed the CSF contamination terms from the equations: Inline graphic from RTOP equation and Inline graphic from MSD equation, to remove the partial volume effect from the computation of these measures.

Therefore, the final equations that we used for computing RTOP and MSD are the following:

graphic file with name M23.gif
graphic file with name M24.gif

References

  1. Almasy L, Blangero J. 1998. Multipoint quantitative-trait linkage analysis in general pedigrees. Am J Hum Genet.  62(5):1198–1211. doi: 10.1086/301844. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Amunts K, Kedo O, Kindler M, Pieperhoff P, Mohlberg H, Shah NJ, Habel U, Schneider F, Zilles K. 2005. Cytoarchitectonic mapping of the human amygdala, hippocampal region and entorhinal cortex: intersubject variability and probability maps. Anat Embryol (Berl).  210(5–6):343–352. doi: 10.1007/s00429-005-0025-5. [DOI] [PubMed] [Google Scholar]
  3. Amunts K, Weiss PH, Mohlberg H, Pieperhoff P, Eickhoff S, Gurd JM, Marshall JC, Shah NJ, Fink GR, Zilles K. 2004. Analysis of neural mechanisms underlying verbal fluency in cytoarchitectonically defined stereotaxic space–the roles of Brodmann areas 44 and 45. NeuroImage.  22(1):42–56. doi: 10.1016/j.neuroimage.2003.12.031. [DOI] [PubMed] [Google Scholar]
  4. Amunts K, Zilles K. 2015. Architectonic mapping of the human brain beyond Brodmann. Neuron.  88(6):1086–1107. doi: 10.1016/j.neuron.2015.12.001. [DOI] [PubMed] [Google Scholar]
  5. Andersen P, Morris R, Amaral D, O’Keefe J, Bliss D, Bliss T. 2007. The hippocampus book. New York: Oxford University Press. [Google Scholar]
  6. Assaf Y, Ben-Bashat D, Chapman J, Peled S, Biton IE, Kafri M, Segev Y, Hendler T, Korczyn AD, Graif M  et al.  2002. High b-value q-space analyzed diffusion-weighted MRI: application to multiple sclerosis. Magn Reson Med.  47(1):115–126. doi: 10.1002/mrm.10040. [DOI] [PubMed] [Google Scholar]
  7. Athanas K, Mauney SL, Woo T-W. 2015. Increased extracellular clusterin in the prefrontal cortex in schizophrenia. Schizophr Res.  169(0):381–385. doi: 10.1016/j.schres.2015.10.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Avram AV, Sarlls JE, Barnett AS, Özarslan E, Thomas C, Irfanoglu MO, Hutchinson E, Pierpaoli C, Basser PJ. 2016. Clinical feasibility of using mean apparent propagator (MAP) MRI to characterize brain tissue microstructure. NeuroImage.  127:422–434. doi: 10.1016/j.neuroimage.2015.11.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Bartley AJ, Jones DW, Weinberger DR. 1997. Genetic variability of human brain size and cortical gyral patterns. Brain J Neurol.  120(Pt 2):257–269. doi: 10.1093/brain/120.2.257 [DOI] [PubMed] [Google Scholar]
  10. Basser PJ, Mattiello J, LeBihan D. 1994. MR diffusion tensor spectroscopy and imaging. Biophys J.  66(1):259–267. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Baumgartner R, Joshi A, Feng D, Zanderigo F, Ogden RT. 2018. Statistical evaluation of test-retest studies in PET brain imaging. EJNMMI Res.  8(1):13. doi: 10.1186/s13550-018-0366-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Bay V, Kjølby BF, Iversen NK, Mikkelsen IK, Ardalan M, Nyengaard JR, Jespersen SN, Drasbek KR, Østergaard L et al.  2018. Stroke infarct volume estimation in fixed tissue: Comparison of diffusion kurtosis imaging to diffusion weighted imaging and histology in a rodent MCAO model. PLoS One.  13(4):e0196161. doi: 10.1371/journal.pone.0196161. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Bekris LM, Yu C-E, Bird TD, Tsuang DW. 2010. Genetics of Alzheimer disease. J Geriatr Psychiatry Neurol.  23(4):213–227. doi: 10.1177/0891988710383571. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Bonnans C, Chou J, Werb Z. 2014. Remodelling the extracellular matrix in development and disease. Nat Rev Mol Cell Biol.  15(12):786–801. doi: 10.1038/nrm3904. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Boscolo Galazzo I, Brusini L, Obertino S, Zucchelli M, Granziera C, Menegaz G. 2018. On the viability of diffusion MRI-based microstructural biomarkers in ischemic stroke. Front Neurosci.  12:92. doi: 10.3389/fnins.2018.00092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Everitt B. 1998. The Cambridge dictionary of statistics. Cambridge, UK: Cambridge University Press. [Google Scholar]
  17. Brown CE. 1998. Coefficient of variation In: Brown CE, editor. Applied multivariate statistics in geohydrology and related sciences. Berlin: Springer; pp. 155–157. [Google Scholar]
  18. Dean K, Murray RM. 2005. Environmental risk factors for psychosis. Dialogues Clin Neurosci.  7(1):69–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Demjaha A, MacCabe JH, Murray RM. 2012. How genes and environmental factors determine the different neurodevelopmental trajectories of schizophrenia and bipolar disorder. Schizophr Bull.  38(2):209–214. doi: 10.1093/schbul/sbr100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Desikan RS, Ségonne F, Fischl B, Quinn BT, Dickerson BC, Blacker D, Buckner RL, Dale AM, Maguire RP, Hyman BT  et al.  2006. An automated labeling system for subdividing the human cerebral cortex on MRI scans into gyral based regions of interest. NeuroImage.  31(3):968–980. doi: 10.1016/j.neuroimage.2006.01.021. [DOI] [PubMed] [Google Scholar]
  21. Dickson DW. 2012. Parkinson’s disease and parkinsonism: neuropathology. Cold Spring Harb Perspect Med.  2(8):a009258. doi: 10.1101/cshperspect.a009258. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Dubois J, Benders M, Borradori-Tolsa C, Cachia A, Lazeyras F, Ha-Vinh Leuchter R, Sizonenko SV, Warfield SK, Mangin JF, Hüppi PS. 2008. Primary cortical folding in the human newborn: an early marker of later functional development. Brain J Neurol.  131(Pt 8):2028–2041. doi: 10.1093/brain/awn137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Duque A, Krsnik Z, Kostović I, Rakic P. 2016. Secondary expansion of the transient subplate zone in the developing cerebrum of human and nonhuman primates. Proc Natl Acad Sci U S A.  113(35):9892–9897. doi: 10.1073/pnas.1610078113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Elman JA, Panizzon MS, Hagler DJ, Fennema-Notestine C, Eyler LT, Gillespie NA, Neale MC, Lyons MJ, Franz CE, McEvoy LK  et al.  2017. Genetic and environmental influences on cortical mean diffusivity. NeuroImage.  146:90–99. doi: 10.1016/j.neuroimage.2016.11.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Fick RHJ, Wassermann D, Caruyer E, Deriche R. 2016. MAPL: tissue microstructure estimation using Laplacian-regularized MAP-MRI and its application to HCP data. NeuroImage.  134:365–385. doi: 10.1016/j.neuroimage.2016.03.046. [DOI] [PubMed] [Google Scholar]
  26. Fischl B, Dale AM. 2000. Measuring the thickness of the human cerebral cortex from magnetic resonance images. Proc Natl Acad Sci U S A.  97(20):11050–11055. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, van der Kouwe A, Killiany R, Kennedy D, Klaveness S et al.  2002. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron.  33:341–355. doi: 10.1016/s0896-6273(02)00569-x. [DOI] [PubMed] [Google Scholar]
  28. Franke B, Stein JL, Ripke S, Anttila V, Hibar DP, Hulzen KJE, Arias-Vasquez A, Smoller JW, Nichols TE, Neale MC  et al.  2016. Genetic influences on schizophrenia and subcortical brain volumes: large-scale proof of concept. Nat Neurosci.  19(3):420–431. doi: 10.1038/nn.4228. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Gejman P, Sanders A, Duan J. 2010. The role of genetics in the etiology of schizophrenia. Psychiatr Clin North Am.  33(1):35–66. doi: 10.1016/j.psc.2009.12.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Glasser MF, Sotiropoulos SN, Wilson JA, Coalson TS, Fischl B, Andersson JL, Xu J, Jbabdi S, Webster M, Polimeni JR  et al.  2013. The minimal preprocessing pipelines for the Human Connectome Project. NeuroImage.  80:105–124. doi: 10.1016/j.neuroimage.2013.04.127. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Grove EA, Tole S. 1999. Patterning events and specification signals in the developing hippocampus. Cereb Cortex.  9(6):551–561. doi: 10.1093/cercor/9.6.551. [DOI] [PubMed] [Google Scholar]
  32. Guen YL, Karkar S, Grigis A, Philippe C, Mangin J, Frouin V. 2019. Heritability of surface area and cortical thickness: a comparison between the Human Connectome Project and the UK Biobank dataset In: 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019), Venice, Italy, 2019. 1887–1890. doi: 10.1109/ISBI.2019.8759539. [DOI]
  33. Gustavson DE, Panizzon MS, Franz CE, Friedman NP, Reynolds CA, Jacobson KC, Xian H, Lyons MJ, Kremen WS. 2018. Genetic and environmental architecture of executive functions in midlife. Neuropsychology.  32(1):18–30. doi: 10.1037/neu0000389. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Guzzoni V, Ribeiro MBT, Lopes GN, Cássia Marqueti R, Andrade RV, Selistre-de-Araujo HS, Durigan JLQ. 2018. Effect of resistance training on extracellular matrix adaptations in skeletal muscle of older rats. Front Physiol.  9:374. doi: 10.3389/fphys.2018.00374. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Hagmann P, Jonasson L, Maeder P, Thiran J-P, Wedeen VJ, Meuli R. 2006. Understanding diffusion MR imaging techniques: from scalar diffusion-weighted imaging to diffusion tensor imaging and beyond. Radiogr Rev Publ Radiol Soc N Am Inc.  26(Suppl 1):S205–S223. doi: 10.1148/rg.26si065510. [DOI] [PubMed] [Google Scholar]
  36. Hendrickson TJ, Mueller BA, Sowell ER, Mattson SN, Coles CD, Kable JA, Jones KL, Boys CJ, Lim KO, Riley EP  et al.  2017. Cortical gyrification is abnormal in children with prenatal alcohol exposure. NeuroImage Clin.  15:391–400. doi: 10.1016/j.nicl.2017.05.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Jansen AG, Mous SE, White T, Posthuma D, Polderman TJC. 2015. What twin studies tell us about the heritability of brain development, morphology, and function: a review. Neuropsychol Rev.  25(1):27–46. doi: 10.1007/s11065-015-9278-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Jensen JH, Helpern JA. 2010. MRI quantification of non-Gaussian water diffusion by kurtosis analysis. NMR Biomed.  23(7):698–710. doi: 10.1002/nbm.1518. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Jones DK, Cercignani M. 2010. Twenty-five pitfalls in the analysis of diffusion MRI data. NMR Biomed.  23(7):803–820. doi: 10.1002/nbm.1543. [DOI] [PubMed] [Google Scholar]
  40. Joshi AA, Leporé N, Joshi SH, Lee AD, Barysheva M, Stein JL, McMahon KL, Johnson K, Zubicaray GI, Martin NG  et al.  2011. The contribution of genes to cortical thickness and volume. Neuroreport.  22(3):101–105. doi: 10.1097/WNR.0b013e3283424c84. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Kärger J, Pfeifer H, Heink W. 1988. Principles and applications of self-diffusion measurements by nuclear magnetic resonance. Adv Magn Reson.  12:1–89. [Google Scholar]
  42. Keller SS, Gerdes JS, Mohammadi S, Kellinghaus C, Kugel H, Deppe K, Ringelstein EB, Evers S, Schwindt W, Deppe M. 2012. Volume estimation of the thalamus using freesurfer and stereology: consistency between methods. Neuroinformatics.  10(4):341–350. doi: 10.1007/s12021-012-9147-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Kochunov P, Donohue B, Mitchell BD, Ganjgahi H, Adhikari B, Ryan M, Medland SE, Jahanshad N, Thompson PM, Blangero J  et al.  2019a. Genomic kinship construction to enhance genetic analyses in the human connectome project data. Hum Brain Mapp.  40(5):1677–1688. doi: 10.1002/hbm.24479. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Kochunov P, Patel B, Ganjgahi H, Donohue B, Ryan M, Hong EL, Chen X, Adhikari B, Jahanshad N, Thompson PM  et al.  2019b. Homogenizing estimates of heritability among SOLAR-eclipse, OpenMx, APACE, and FPHI software packages in neuroimaging data. Front Neuroinform.  13:16. doi: 10.3389/fninf.2019.00016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Koran ME, Thornton-Wells TA, Jahanshad N, Glahn DC, Thompson PM, Blangero J, Nichols TE, Kochunov P, Landman BA. 2014. Impact of family structure and common environment on heritability estimation for neuroimaging genetics studies using sequential oligogenic linkage analysis routines. J Med Imaging.  1(1):014005. doi: 10.1117/1.JMI.1.1.014005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Kramer AF, Bherer L, Colcombe SJ, Dong W, Greenough WT. 2004. Environmental influences on cognitive and brain plasticity during aging. J Gerontol A Biol Sci Med Sci.  59(9):M940–M957. doi: 10.1093/gerona/59.9.m940. [DOI] [PubMed] [Google Scholar]
  47. Kremen WS, Panizzon MS, Neale MC, Fennema-Notestine C, Prom-Wormley E, Eyler LT, Stevens A, Franz CE, Lyons MJ, Grant MD  et al.  2012. Heritability of brain ventricle volume: converging evidence from inconsistent results. Neurobiol Aging.  33(1):1–8. doi: 10.1016/j.neurobiolaging.2010.02.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Kroenke CD, Bayly PV. 2018. How forces fold the cerebral cortex. J Neurosci.  38(4):767–775. doi: 10.1523/JNEUROSCI.1105-17.2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Landrigan PJ, Sonawane B, Butler RN, Trasande L, Callan R, Droller D. 2005. Early environmental origins of neurodegenerative disease in later life. Environ Health Perspect.  113(9):1230–1233. doi: 10.1289/ehp.7571. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Le Bihan D, Iima M. 2015. Diffusion magnetic resonance imaging: what water tells us about biological tissues. PLoS Biol.  13(7):e1002203. doi: 10.1371/journal.pbio.1002203. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Llinares-Benadero C, Borrell V. 2019. Deconstructing cortical folding: genetic, cellular and mechanical determinants. Nat Rev Neurosci.  20(3):161–176. doi: 10.1038/s41583-018-0112-2. [DOI] [PubMed] [Google Scholar]
  52. Lovie P. 2005. Coefficient of variation In: Encyclopedia of statistics in behavioral science. American Cancer Society. Hoboken, NJ: John Wiley & Sons, Ltd. [Google Scholar]
  53. Lukies MW, Watanabe Y, Tanaka H, Takahashi H, Ogata S, Omura K, Yorifuji S, Tomiyama N. 2017. Heritability of brain volume on MRI in middle to advanced age: a twin study of Japanese adults. PLoS One.  12(4):e0175800. doi: 10.1371/journal.pone.0175800. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. McClelland JL, McNaughton BL, O'Reilly RC. 1995. Why there are complementary learning systems in the hippocampus and neocortex: insights from the successes and failures of connectionist models of learning and memory. Psychol Rev.  102(3):419–457. doi: 10.1037/0033-295X.102.3.419. [DOI] [PubMed] [Google Scholar]
  55. McEwen BS, Getz L. 2013. Lifetime experiences, the brain and personalized medicine: an integrative perspective. Metabolism.  62:S20–S26. doi: 10.1016/j.metabol.2012.08.020. [DOI] [PubMed] [Google Scholar]
  56. McEwen BS, Nasca C, Gray JD. 2016. Stress effects on neuronal structure: hippocampus, amygdala, and prefrontal cortex. Neuropsychopharmacology.  41(1):3–23. doi: 10.1038/npp.2015.171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Mitra PP, Halperin BI. 1995. Effects of finite gradient-pulse widths in pulsed-field-gradient diffusion measurements. J Magn Reson A.  113(1):94–101. doi: 10.1006/jmra.1995.1060. [DOI] [Google Scholar]
  58. Mulkern RV, Gudbjartsson H, Westin CF, Zengingonul HP, Gartner W, Guttmann CR, Robertson RL, Kyriakos W, Schwartz R, Holtzman D  et al.  1999. Multi-component apparent diffusion coefficients in human brain. NMR Biomed.  12(1):51–62. [DOI] [PubMed] [Google Scholar]
  59. Nazeri A, Mulsant BH, Rajji TK, Levesque ML, Pipitone J, Stefanik L, Shahab S, Roostaei T, Wheeler AL, Chavez S  et al.  2017. Gray matter neuritic microstructure deficits in schizophrenia and bipolar disorder. Biol Psychiatry.  82(10):726–736. doi: 10.1016/j.biopsych.2016.12.005. [DOI] [PubMed] [Google Scholar]
  60. Ning L, Westin C-F, Rathi Y. 2015. Estimating diffusion propagator and its moments using directional radial basis functions. IEEE Trans Med Imaging.  34(10):2058–2078. doi: 10.1109/TMI.2015.2418674. [DOI] [PMC free article] [PubMed] [Google Scholar]
  61. Norman KA. 2010. How hippocampus and cortex contribute to recognition memory: revisiting the complementary learning systems model. Hippocampus.  20(11):1217–1227. doi: 10.1002/hipo.20855. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. O’Donnell LJ, Westin C-F. 2011. An introduction to diffusion tensor image analysis. Neurosurg Clin N Am.  22(2):185–viii. doi: 10.1016/j.nec.2010.12.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Oldham MC, Horvath S, Geschwind DH. 2006. Conservation and evolution of gene coexpression networks in human and chimpanzee brains. Proc Natl Acad Sci U S A.  103(47):17973–17978. doi: 10.1073/pnas.0605938103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Özarslan E, Koay CG, Shepherd TM, Komlosh ME, İrfanoğlu MO, Pierpaoli C, Basser PJ. 2013. Mean apparent propagator (MAP) MRI: a novel diffusion imaging method for mapping tissue microstructure. NeuroImage.  78:16–32. doi: 10.1016/j.neuroimage.2013.04.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Pengas G, Pereira JMS, Williams GB, Nestor PJ. 2009. Comparative reliability of total intracranial volume estimation methods and the influence of atrophy in a longitudinal semantic dementia cohort. J Neuroimaging.  19(1):37–46. doi: 10.1111/j.1552-6569.2008.00246.x. [DOI] [PubMed] [Google Scholar]
  66. Peper JS, Brouwer RM, Boomsma DI, Kahn RS, Hulshoff Pol HE. 2007. Genetic influences on human brain structure: a review of brain imaging studies in twins. Hum Brain Mapp.  28(6):464–473. doi: 10.1002/hbm.20398. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Pines AR, Cieslak M, Larsen B, Baum GL, Cook PA, Adebimpe A, Dávila DG, Elliott MA, Jirsaraie R, Murtha K  et al.  2020. Leveraging multi-shell diffusion for studies of brain development in youth and young adulthood. Dev Cogn Neurosci.  43:100788. doi: 10.1016/j.dcn.2020.100788. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Quezada S, Castillo-Melendez M, Walker DW, Tolcos M. 2018. Development of the cerebral cortex and the effect of the intrauterine environment. J Physiol.  596(23):5665–5674. doi: 10.1113/JP277151. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Rajkowska G, Stockmeier CA. 2013. Astrocyte pathology in major depressive disorder: insights from human postmortem brain tissue. Curr Drug Targets.  14(11):1225–1236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Rakic P. 1988. Specification of cerebral cortical areas. Sci Wash.  241(4862):170. [DOI] [PubMed] [Google Scholar]
  71. Rakic P, Swaab DF. 1988. Defects of neuronal migration and the pathogenesis of cortical malformations In: Boer GJ, Feenstra MGP, Mirmiran M, Swaab DF, Haaren F, editors. Progress in brain research. Elsevier. (Biochemical Basis of Functional Neuroteratology), 73:15–37. doi: 10.1016/s0079-6123(08)60494-x. [DOI] [PubMed] [Google Scholar]
  72. Rathi Y, Gagoski B, Setsompop K, Michailovich O, Grant PE, Westin C-F. 2013. Diffusion propagator estimation from sparse measurements in a tractography framework. Med Image Comput Comput-Assist Interv MICCAI Int Conf Med Image Comput Comput-Assist Interv.  16(3):510–517. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Rathi Y, Pasternak O, Savadjiev P, Michailovich O, Bouix S, Kubicki M, Westin C-F, Makris N, Shenton ME. 2014. Gray matter alterations in early aging: a diffusion magnetic resonance imaging study. Hum Brain Mapp.  35(8):3841–3856. doi: 10.1002/hbm.22441. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Reas ET, Hagler DJ, White NS, Kuperman JM, Bartsch H, Wierenga CE, Galasko D, Brewer JB, Dale AM, McEvoy LK. 2018. Microstructural brain changes track cognitive decline in mild cognitive impairment. NeuroImage Clin.  20:883–891. doi: 10.1016/j.nicl.2018.09.027. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. Rogers J, Kochunov P, Zilles K, Shelledy W, Lancaster J, Thompson P, Duggirala R, Blangero J, Fox PT, Glahn DC. 2010. On the genetic architecture of cortical folding and brain volume in primates. NeuroImage.  53(3):1103–1108. doi: 10.1016/j.neuroimage.2010.02.020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Schmitt JE, Eyler LT, Giedd JN, Kremen WS, Kendler KS, Neale MC. 2007. Review of twin and family studies on neuroanatomic phenotypes and typical neurodevelopment. Twin Res Hum Genet Off J Int Soc Twin Stud.  10(5):683–694. doi: 10.1375/twin.10.5.683. [DOI] [PMC free article] [PubMed] [Google Scholar]
  77. Seitz J, Rathi Y, Lyall A, Pasternak O, Del Re EC, Niznikiewicz M, Nestor P, Seidman LJ, Petryshen TL, Mesholam-Gately RI  et al.  2018. Alteration of gray matter microstructure in schizophrenia. Brain Imaging Behav.  12(1):54–63. doi: 10.1007/s11682-016-9666-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Shadrina M, Bondarenko EA, Slominsky PA. 2018. Genetics factors in major depression disease. Front Psychiatry.  9:334. doi: 10.3389/fpsyt.2018.00334. [DOI] [PMC free article] [PubMed] [Google Scholar]
  79. Shahim P, Holleran L, Kim JH, Brody DL. 2017. Test-retest reliability of high spatial resolution diffusion tensor and diffusion kurtosis imaging. Sci Rep.  7(1):1–14. doi: 10.1038/s41598-017-11747-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Shankle WR, Romney AK, Landing BH, Hara J. 1998. Developmental patterns in the cytoarchitecture of the human cerebral cortex from birth to 6 years examined by correspondence analysis. Proc Natl Acad Sci U S A.  95(7):4023–4028. doi: 10.1073/pnas.95.7.4023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Shaw P, Kabani NJ, Lerch JP, Eckstrand K, Lenroot R, Gogtay N, Greenstein D, Clasen L, Evans A, Rapoport JL  et al.  2008. Neurodevelopmental trajectories of the human cerebral cortex. J Neurosci.  28(14):3586–3594. doi: 10.1523/JNEUROSCI.5309-07.2008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Sotiropoulos SN, Jbabdi S, Xu J, Andersson JL, Moeller S, Auerbach EJ, Glasser MF, Hernandez M, Sapiro G, Jenkinson M  et al.  2013. Advances in diffusion MRI acquisition and processing in the Human Connectome Project. NeuroImage.  80:125–143. doi: 10.1016/j.neuroimage.2013.05.057. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Strominger NL, Demarest RJ, Laemle LB. 2012. Cerebral cortex In: Strominger NL, Demarest RJ, Laemle LB, editors. Noback’s human nervous system: structure and function. 7th ed Totowa (NJ): Humana Press, pp. 429–451. [Google Scholar]
  84. Tanner CM, Ottman R, Goldman SM, Ellenberg J, Chan P, Mayeux R, Langston JW. 1999. Parkinson disease in twins: an etiologic study. JAMA.  281(4):341–346. 10.1001/jama.281.4.341. [DOI] [PubMed] [Google Scholar]
  85. R Core Team 2012. 2018. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing, Httpwww R-Proj Org. [Google Scholar]
  86. Tenesa A, Haley CS. 2013. The heritability of human disease: estimation, uses and abuses. Nat Rev Genet.  14(2):139–149. doi: 10.1038/nrg3377. [DOI] [PubMed] [Google Scholar]
  87. Thompson PM, Cannon TD, Narr KL, van ET, Poutanen V-P, Huttunen M, Lönnqvist J, Standertskjöld-Nordenstam C-G, Kaprio J, Khaledy M  et al.  2001. Genetic influences on brain structure. Nat Neurosci.  4(12):1253–1258. doi: 10.1038/nn758. [DOI] [PubMed] [Google Scholar]
  88. Timmons JA, Jansson E, Fischer H, Gustafsson T, Greenhaff PL, Ridden J, Rachman J, Sundberg CJ. 2005. Modulation of extracellular matrix genes reflects the magnitude of physiological adaptation to aerobic exercise training in humans. BMC Biol.  3(1):19. doi: 10.1186/1741-7007-3-19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  89. Tucker-Drob EM, Briley DA, Harden KP. 2013. Genetic and environmental influences on cognition across development and context. Curr Dir Psychol Sci.  22(5):349–355. doi: 10.1177/0963721413485087. [DOI] [PMC free article] [PubMed] [Google Scholar]
  90. Uğurbil K, Xu J, Auerbach EJ, Moeller S, Vu AT, Duarte-Carvajalino JM, Lenglet C, Wu X, Schmitter S, Van de Moortele PF  et al.  2013. Pushing spatial and temporal resolution for functional and diffusion MRI in the human connectome project. NeuroImage.  80:80–104. doi: 10.1016/j.neuroimage.2013.05.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  91. Van Essen DC, Smith SM, Barch DM, Behrens Timothy EJ, Yacoub E, Ugurbil K. 2013. The WU-Minn Human Connectome Project: an overview. NeuroImage.  80:62–79. doi: 10.1016/j.neuroimage.2013.05.041. [DOI] [PMC free article] [PubMed] [Google Scholar]
  92. Voss MW, Prakash RS, Erickson KI, Basak C, Chaddock L, Kim JS, Alves H, Heo S, Szabo AN, White SM  et al.  2010. Plasticity of brain networks in a randomized intervention trial of exercise training in older adults. Front Aging Neurosci.  2:32. doi: 10.3389/fnagi.2010.00032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  93. Weston PSJ, Simpson IJA, Ryan NS, Ourselin S, Fox NC. 2015. Diffusion imaging changes in grey matter in Alzheimer’s disease: a potential marker of early neurodegeneration. Alzheimers Res Ther.  7(1):47. doi: 10.1186/s13195-015-0132-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  94. Wu W, McAnulty G, Hamoda HM, Sarill K, Karmacharya S, Gagoski B, Ning L, Grant PE, Shenton ME, Waber DP  et al.  2019. Detecting microstructural white matter abnormalities of frontal pathways in children with ADHD using advanced diffusion models. Brain Imaging Behav. doi: 10.1007/s11682-019-00108-5. [DOI] [PubMed] [Google Scholar]
  95. Zhang Xin CX, Xiaobo Z. 2018. The impact of exposure to air pollution on cognitive performance. Proc Natl Acad Sci U S A.  115(37):9193–9197. doi: 10.1073/pnas.1809474115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  96. Zhang Y, Inder TE, Neil JJ, Dierker DL, Alexopoulos D, Anderson PJ, Van Essen DC. 2015. Cortical structural abnormalities in very preterm children at 7 years of age. NeuroImage.  109:469–479. doi: 10.1016/j.neuroimage.2015.01.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  97. Zilles K, Palomero-Gallagher N, Amunts K. 2013. Development of cortical folding during evolution and ontogeny. Trends Neurosci.  36(5):275–284. doi: 10.1016/j.tins.2013.01.006. [DOI] [PubMed] [Google Scholar]

Articles from Cerebral Cortex (New York, NY) are provided here courtesy of Oxford University Press

RESOURCES