Short superficial white matter and aging: A longitudinal multi-site study of 1293 subjects and 2711 sessions

Kurt G. Schilling, Derek B. Archer, Fang‐Cheng Yeh, François Rheault, Leon Y. Cai, Andrea T. Shafer, Susan M. Resnick, Timothy J. Hohman, Angela L. Jefferson, Adam W. Anderson, Hakmook Kang, Bennett A. LandmanAging 2023 · DOI

TL;DR

This study leverages advanced tractography techniques on a large, multi-site dataset to characterize how superficial white matter (SWM) changes with normal aging. They found that while microstructural changes in SWM mirror those in deeper white matter (e.g., decreased FA, increased diffusivity), its macrostructural changes (volume and length) are less pronounced and more heterogeneous, offering a new perspective on brain aging.

What Problem Does This Paper Address?

The brain's white matter is crucial for communication between different regions, and its integrity is known to decline with age. While much research has focused on the major, long-range white matter pathways (association, projection, and commissural fibers), the superficial white matter (SWM) – the network of short fibers just beneath the cortex – has been relatively understudied. This is partly due to technical challenges in accurately mapping these intricate, short fibers using diffusion MRI tractography. Given that SWM constitutes a significant portion of the brain's white matter volume and plays a role in brain function and plasticity, understanding its aging trajectory is essential for a complete picture of brain aging and for distinguishing normal aging from age-related neurodegenerative diseases.

Methods

The researchers employed a sophisticated tractography pipeline to dissect SWM. Key steps included:

  • Preprocessing: Standard diffusion MRI preprocessing (motion correction, eddy current correction, susceptibility distortion correction).
  • Fiber Orientation Estimation: Using multi-tissue constrained spherical deconvolution (CSD) to model complex fiber orientations.
  • Anatomically Constrained Tractography (ACT): Seeding streamlines from the white/gray matter boundary and forcing them to terminate at the neocortex to specifically target SWM. This, combined with a maximum streamline length of 40mm, ensured the focus on "short association fibers."
  • Bundle Segmentation: Utilizing the Desikan-Killiany atlas to assign streamlines to connection matrices, creating potential SWM bundles.
  • Filtering: Applying QuickBundles clustering and a reproducibility threshold (95% of the population) to select robust SWM bundles, resulting in 132 bundles.
  • Feature Extraction: Quantifying six features for each bundle: fractional anisotropy (FA), mean diffusivity (MD), radial diffusivity (RD), axial diffusivity (AD) for microstructural properties, and length and volume for macrostructural properties.
  • Statistical Analysis: Employing linear mixed-effects modeling to assess the association between age and each feature, accounting for dataset differences and individual variability.

Datasets

The study utilized data from three large, well-established cohorts:

  • Baltimore Longitudinal Study of Aging (BLSA): A longitudinal cohort with 741 participants and 2711 sessions.
  • Vanderbilt Memory and Aging Project (VMAP): A longitudinal cohort with 187 participants and 558 sessions.
  • Cambridge Centre for Ageing and Neuroscience (Cam-CAN): A cross-sectional cohort with 356 participants and 365 sessions.

All participants were aged 50 and above and had no diagnosed cognitive impairment at baseline.

Key Insights and Ideas

  • SWM Microstructure Mirrors Deep White Matter Aging: The study found that microstructural changes in SWM (FA decreasing, diffusivities increasing with age) are remarkably similar to those observed in larger, long-range white matter pathways. This suggests a common underlying biological mechanism for microstructural aging across different white matter types.
  • Macrostructural Changes are More Subtle and Heterogeneous: Unlike the more consistent declines seen in deep white matter, SWM volume and length showed more varied patterns of change across different brain regions and between inter-gyri (connecting adjacent gyri) and intra-gyri (within a gyrus) fibers.
  • Frontal and Temporal Lobes are Key Areas of Change: Age-related microstructural changes in SWM were most pronounced in the frontal, temporal, and temporoparietal regions, highlighting these areas as particularly vulnerable to aging processes.
  • Inter- vs. Intra-Gyri Differences: Intra-gyral SWM showed a greater decrease in volume with age compared to inter-gyral SWM. This difference is hypothesized to be related to changes in cortical folding (sulcal widening) and thinning cortex.
  • SWM as a Complementary Aging Biomarker: The findings suggest that SWM, particularly its macrostructural features, could provide unique insights into brain aging that complement information from gray matter and deep white matter.

Key Results

  • Microstructure: Across all SWM pathways, FA showed a significant negative association with age, while AD, RD, and MD showed significant positive associations with age. These associations were strongest in frontal, temporal, and temporoparietal regions.
  • Macrostructure: SWM length generally decreased with age, though with less pronounced effects than microstructural changes. SWM volume changes were more heterogeneous, with intra-gyral SWM showing larger decreases than inter-gyral SWM.
  • Comparison to Long-Range Pathways: While SWM microstructural aging patterns were similar to long-range pathways, SWM's length and volume changes were less pronounced than those observed in deeper white matter.

Limitations and Future Work

  • Partial Volume Effects: The authors acknowledge that tractography is susceptible to partial volume effects, where signals from adjacent tissues (gray matter, other white matter tracts) can influence the measured indices. This is particularly relevant for SWM, which is closely associated with the cortex.
  • Resolution and Angular Information: While the study used advanced methods, the resolution and angular information of the diffusion MRI data, while sufficient for previous validation, could be improved in future studies to further reduce variability.
  • Non-linear Aging: The study focused on participants aged 50+, and while linear mixed-effects modeling was used, future work could explore non-linear aging trajectories across the entire lifespan.
  • Methodological Variations: Different tractography and segmentation methods exist, and while the chosen pipeline is robust, results might vary slightly with alternative approaches.
  • Future Directions: The authors suggest integrating SWM analysis with gray matter and long-range pathway data for a more comprehensive understanding of brain aging. They also highlight the potential of SWM analysis for understanding cognitive decline and identifying networks susceptible to disease, and recommend future studies incorporate similar advanced tractography techniques with similar datasets.

Summary updated

Abstract

It is estimated that short association fibers running immediately beneath the cortex may make up as much as 60% of the total white matter volume. However, these have been understudied relative to the long-range association, projection, and commissural fibers of the brain. This is largely because of limitations of diffusion MRI fiber tractography, which is the primary methodology used to non-invasively study the white matter connections. Inspired by recent anatomical considerations and methodological improvements in superficial white matter (SWM) tractography, we aim to characterize changes in these fiber systems in cognitively normal aging, which provide insight into the biological foundation of age-related cognitive changes, and a better understanding of how age-related pathology differs from healthy aging. To do this, we used three large, longitudinal and cross-sectional datasets (N = 1293 subjects, 2711 sessions) to quantify microstructural features and length/volume features of several SWM systems. We find that axial, radial, and mean diffusivities show positive associations with age, while fractional anisotropy has negative associations with age in SWM throughout the entire brain. These associations were most pronounced in the frontal, temporal, and temporoparietal regions. Moreover, measures of SWM volume and length decrease with age in a heterogenous manner across the brain, with different rates of change in inter-gyri and intra-gyri SWM, and at slower rates than well-studied long-range white matter pathways. These features, and their variations with age, provide the background for characterizing normal aging, and, in combination with larger association pathways and gray matter microstructural features, may provide insight into fundamental mechanisms associated with aging and cognition.
BibTeX
@inproceedings{schilling2023short,
  author = {Kurt G. Schilling and Derek B. Archer and Fang‐Cheng Yeh and François Rheault and Leon Y. Cai and Andrea T. Shafer and Susan M. Resnick and Timothy J. Hohman and Angela L. Jefferson and Adam W. Anderson and Hakmook Kang and Bennett A. Landman},
  title = {{Short superficial white matter and aging: A longitudinal multi-site study of 1293 subjects and 2711 sessions}},
  booktitle = {Aging},
  year = {2023},
  doi = {10.1016/j.nbas.2023.100067},
}
Figure