Limits...
Synergistic Effects of Age on Patterns of White and Gray Matter Volume across Childhood and Adolescence(1,2,3).

Bray S, Krongold M, Cooper C, Lebel C - eNeuro (2015)

Bottom Line: Linear effects of age on white and gray matter volume were modeled within four age bins, spanning 4-18 years, each including 90 participants (45 male).Four white matter clusters were identified, each with a dominant direction of underlying fibers: anterior-posterior, left-right, and two clusters with superior-inferior directions.Pairs of gray and white matter clusters followed parallel slope trajectories, with white matter changes generally positive from 8 years onward (indicating volume increases) and gray matter negative (decreases).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Radiology, Cumming School of Medicine, University of Calgary , Calgary, Alberta, Canada T2N 1N4 ; Department of Pediatrics, Cumming School of Medicine, University of Calgary , Calgary, Alberta, Canada T2N 1N4 ; Child and Adolescent Imaging Research Program, Alberta Children's Hospital , Calgary, Alberta, Canada T3B 6A8 ; Alberta Children's Hospital Research Institute , Calgary, Alberta, Canada T3B 6A8.

ABSTRACT
The human brain develops with a nonlinear contraction of gray matter across late childhood and adolescence with a concomitant increase in white matter volume. Across the adult population, properties of cortical gray matter covary within networks that may represent organizational units for development and degeneration. Although gray matter covariance may be strongest within structurally connected networks, the relationship to volume changes in white matter remains poorly characterized. In the present study we examined age-related trends in white and gray matter volume using T1-weighted MR images from 360 human participants from the NIH MRI study of Normal Brain Development. Images were processed through a voxel-based morphometry pipeline. Linear effects of age on white and gray matter volume were modeled within four age bins, spanning 4-18 years, each including 90 participants (45 male). White and gray matter age-slope maps were separately entered into k-means clustering to identify regions with similar age-related variability across the four age bins. Four white matter clusters were identified, each with a dominant direction of underlying fibers: anterior-posterior, left-right, and two clusters with superior-inferior directions. Corresponding, spatially proximal, gray matter clusters encompassed largely cerebellar, fronto-insular, posterior, and sensorimotor regions, respectively. Pairs of gray and white matter clusters followed parallel slope trajectories, with white matter changes generally positive from 8 years onward (indicating volume increases) and gray matter negative (decreases). As developmental disorders likely target networks rather than individual regions, characterizing typical coordination of white and gray matter development can provide a normative benchmark for understanding atypical development.

No MeSH data available.


Related in: MedlinePlus

Effects of site and resolution on regional volume and clustering results. a, b, Results of an F-contrast for effects of site, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Site effects were identified around the putamen and internal capsule as well as medial orbital and peripheral gray matter. c, d, Results of an F-contrast for effects of resolution, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Effects were again concentrated around the internal capsule, with gray matter effects in dorsal prefrontal, occipital cortices, and cerebellum. e, f, Illustrate clustering results for age β-estimates from gray (e) and white (f) matter models that included effects of site and resolution (left and superior) and from models that did not include these effects (right and inferior). We note that these are largely similar. g, h, Illustrate regions of overlap (purple) and difference (red and blue) in cluster assignment for gray (g) and white (h) matter clusters when site and resolution are taken into account. The only substantial differences in clustering results were in posterior putamen and near the internal capsule. GM, gray matter; WM, white matter.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4596017&req=5

Figure 6: Effects of site and resolution on regional volume and clustering results. a, b, Results of an F-contrast for effects of site, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Site effects were identified around the putamen and internal capsule as well as medial orbital and peripheral gray matter. c, d, Results of an F-contrast for effects of resolution, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Effects were again concentrated around the internal capsule, with gray matter effects in dorsal prefrontal, occipital cortices, and cerebellum. e, f, Illustrate clustering results for age β-estimates from gray (e) and white (f) matter models that included effects of site and resolution (left and superior) and from models that did not include these effects (right and inferior). We note that these are largely similar. g, h, Illustrate regions of overlap (purple) and difference (red and blue) in cluster assignment for gray (g) and white (h) matter clusters when site and resolution are taken into account. The only substantial differences in clustering results were in posterior putamen and near the internal capsule. GM, gray matter; WM, white matter.

Mentions: Previous studies have noted that VBM estimates of gray and white matter volume are sensitive to differences in MR scanner and image resolution (Pardoe et al., 2008; Pereira et al., 2008; Focke et al., 2011; Takao et al., 2013). As the present study made use of a multisite dataset, additional analyses were run to estimate potential impact of these factors on our results. A general linear model was estimated using the entire sample of 360 participants, including a linear effect of resolution and separate regressors modeling effects of each site. F-contrasts were used to identify regions sensitive to these effects. Results of F-contrasts for site are shown in Figure 6, a and b, thresholded at p < 0.001 uncorrected, for gray and white matter, respectively. We observe significant effects of site around the posterior putamen, orbitofrontal, inferior temporal, and peripheral gray matter in Figure 6a. Significant effects on white matter volume were most prominent around the internal capsule (Fig. 6b). Results of F-contrasts for resolution are shown in Figure 6, c and d, thresholded at p < 0.001 uncorrected, for gray and white matter, respectively. Affected gray matter regions were similarly concentrated around the posterior putamen and insula and occipital and dorsal prefrontal regions (Fig. 6c). For white matter, similar to effects of site, effects of resolution were largely concentrated around the internal capsule (Fig. 6d). We next compared clustering results for age β-values from models that did and did not include effects of site and resolution. These results are shown in Figure 6, e and f. We note that clustering results were largely similar between these two models. Figure 6, g and h, shows gray and white matter clusters obtained from these two models overlaid; regions of overlap are shown in purple. The only regions where cluster assignment substantially differed were around the putamen and internal capsule. Overall these results suggest that effects of site and resolution may have a fairly localized effect in subcortical regions, and we note that reliability of cluster assignment in these regions is a limitation of the present work.


Synergistic Effects of Age on Patterns of White and Gray Matter Volume across Childhood and Adolescence(1,2,3).

Bray S, Krongold M, Cooper C, Lebel C - eNeuro (2015)

Effects of site and resolution on regional volume and clustering results. a, b, Results of an F-contrast for effects of site, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Site effects were identified around the putamen and internal capsule as well as medial orbital and peripheral gray matter. c, d, Results of an F-contrast for effects of resolution, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Effects were again concentrated around the internal capsule, with gray matter effects in dorsal prefrontal, occipital cortices, and cerebellum. e, f, Illustrate clustering results for age β-estimates from gray (e) and white (f) matter models that included effects of site and resolution (left and superior) and from models that did not include these effects (right and inferior). We note that these are largely similar. g, h, Illustrate regions of overlap (purple) and difference (red and blue) in cluster assignment for gray (g) and white (h) matter clusters when site and resolution are taken into account. The only substantial differences in clustering results were in posterior putamen and near the internal capsule. GM, gray matter; WM, white matter.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4596017&req=5

Figure 6: Effects of site and resolution on regional volume and clustering results. a, b, Results of an F-contrast for effects of site, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Site effects were identified around the putamen and internal capsule as well as medial orbital and peripheral gray matter. c, d, Results of an F-contrast for effects of resolution, thresholded at p < 0.001 uncorrected, on gray and white matter volume, respectively. Effects were again concentrated around the internal capsule, with gray matter effects in dorsal prefrontal, occipital cortices, and cerebellum. e, f, Illustrate clustering results for age β-estimates from gray (e) and white (f) matter models that included effects of site and resolution (left and superior) and from models that did not include these effects (right and inferior). We note that these are largely similar. g, h, Illustrate regions of overlap (purple) and difference (red and blue) in cluster assignment for gray (g) and white (h) matter clusters when site and resolution are taken into account. The only substantial differences in clustering results were in posterior putamen and near the internal capsule. GM, gray matter; WM, white matter.
Mentions: Previous studies have noted that VBM estimates of gray and white matter volume are sensitive to differences in MR scanner and image resolution (Pardoe et al., 2008; Pereira et al., 2008; Focke et al., 2011; Takao et al., 2013). As the present study made use of a multisite dataset, additional analyses were run to estimate potential impact of these factors on our results. A general linear model was estimated using the entire sample of 360 participants, including a linear effect of resolution and separate regressors modeling effects of each site. F-contrasts were used to identify regions sensitive to these effects. Results of F-contrasts for site are shown in Figure 6, a and b, thresholded at p < 0.001 uncorrected, for gray and white matter, respectively. We observe significant effects of site around the posterior putamen, orbitofrontal, inferior temporal, and peripheral gray matter in Figure 6a. Significant effects on white matter volume were most prominent around the internal capsule (Fig. 6b). Results of F-contrasts for resolution are shown in Figure 6, c and d, thresholded at p < 0.001 uncorrected, for gray and white matter, respectively. Affected gray matter regions were similarly concentrated around the posterior putamen and insula and occipital and dorsal prefrontal regions (Fig. 6c). For white matter, similar to effects of site, effects of resolution were largely concentrated around the internal capsule (Fig. 6d). We next compared clustering results for age β-values from models that did and did not include effects of site and resolution. These results are shown in Figure 6, e and f. We note that clustering results were largely similar between these two models. Figure 6, g and h, shows gray and white matter clusters obtained from these two models overlaid; regions of overlap are shown in purple. The only regions where cluster assignment substantially differed were around the putamen and internal capsule. Overall these results suggest that effects of site and resolution may have a fairly localized effect in subcortical regions, and we note that reliability of cluster assignment in these regions is a limitation of the present work.

Bottom Line: Linear effects of age on white and gray matter volume were modeled within four age bins, spanning 4-18 years, each including 90 participants (45 male).Four white matter clusters were identified, each with a dominant direction of underlying fibers: anterior-posterior, left-right, and two clusters with superior-inferior directions.Pairs of gray and white matter clusters followed parallel slope trajectories, with white matter changes generally positive from 8 years onward (indicating volume increases) and gray matter negative (decreases).

View Article: PubMed Central - HTML - PubMed

Affiliation: Department of Radiology, Cumming School of Medicine, University of Calgary , Calgary, Alberta, Canada T2N 1N4 ; Department of Pediatrics, Cumming School of Medicine, University of Calgary , Calgary, Alberta, Canada T2N 1N4 ; Child and Adolescent Imaging Research Program, Alberta Children's Hospital , Calgary, Alberta, Canada T3B 6A8 ; Alberta Children's Hospital Research Institute , Calgary, Alberta, Canada T3B 6A8.

ABSTRACT
The human brain develops with a nonlinear contraction of gray matter across late childhood and adolescence with a concomitant increase in white matter volume. Across the adult population, properties of cortical gray matter covary within networks that may represent organizational units for development and degeneration. Although gray matter covariance may be strongest within structurally connected networks, the relationship to volume changes in white matter remains poorly characterized. In the present study we examined age-related trends in white and gray matter volume using T1-weighted MR images from 360 human participants from the NIH MRI study of Normal Brain Development. Images were processed through a voxel-based morphometry pipeline. Linear effects of age on white and gray matter volume were modeled within four age bins, spanning 4-18 years, each including 90 participants (45 male). White and gray matter age-slope maps were separately entered into k-means clustering to identify regions with similar age-related variability across the four age bins. Four white matter clusters were identified, each with a dominant direction of underlying fibers: anterior-posterior, left-right, and two clusters with superior-inferior directions. Corresponding, spatially proximal, gray matter clusters encompassed largely cerebellar, fronto-insular, posterior, and sensorimotor regions, respectively. Pairs of gray and white matter clusters followed parallel slope trajectories, with white matter changes generally positive from 8 years onward (indicating volume increases) and gray matter negative (decreases). As developmental disorders likely target networks rather than individual regions, characterizing typical coordination of white and gray matter development can provide a normative benchmark for understanding atypical development.

No MeSH data available.


Related in: MedlinePlus