Sex and APOE ε4 allele differences in longitudinal white matter microstructure in multiple cohorts of aging and Alzheimer's disease

Amalia Peterson, Aditi Sathe, Dimitrios Zaras, Yisu Yang, Alaina Durant, Kacie Deters, Niranjana Shashikumar, Kimberly R. Pechman, Michael E. Kim, Chenyu Gao, Nazirah Mohd Khairi, Zhiyuan Li, Tianyuan Yao, Yuankai Huo, Logan Dumitrescu, Katherine A. Gifford, Jo Ellen Wilson, Francis E. Cambronero, Shannon L. Risacher, Lori L. Beason‐Held, Yang An, Konstantinos Arfanakis, Guray Erus, Christos Davatzikos, Duygu Tosun, Arthur W. Toga, Paul M. Thompson, Elizabeth C. Mormino, Mohamad Habes, Di Wang, Panpan Zhang, Kurt G. Schilling, Marilyn Albert, Walter A. Kukull, Sarah Biber, Bennett A. Landman, Sterling C. Johnson, Julie A. Schneider, Lisa L. Barnes, David A. Bennett, Angela L. Jefferson, Susan M. Resnick, Andrew J. Saykin, Timothy J. Hohman, Derek B. ArcherAlzheimer 2024 · DOI

Abstract: There are prominent differences in white matter microstructure by sex, race, and APOE-ε4 carrier status. This work adds to our understanding of disparities in Alzheimer’s disease (AD). Additional work to understand the etiology of these differences is warranted.

Introduction

As the global burden of Alzheimer’s disease (AD) grows, understanding the biological pathways that lead to cognitive decline is paramount. While it is well-established that females, non-Hispanic Black adults, and carriers of the APOE-ε4 allele face a higher risk of AD, the specific impact of these factors on the brain’s white matter—the critical bundles of myelinated axons connecting brain regions—has remained unclear. This study provides a large-scale longitudinal analysis of white matter microstructure, revealing that these risk factors leave distinct biological signatures on the brain's "wiring."

Addressing the "Partial Volume" Challenge

A significant technical hurdle in neuroimaging is the "partial volume effect." Standard Diffusion MRI (dMRI) often struggles to distinguish between the integrity of actual nerve tissue and the fluid surrounding it. If a scan shows low integrity, it is difficult to determine if the nerves themselves are degrading or if there is simply an increase in extracellular fluid due to inflammation or atrophy.

To solve this, researchers employed advanced "free-water (FW) correction," using a two-compartment model to separate the signal into two distinct metrics:

  1. FW (Free-Water): Representing extracellular water, often linked to neuroinflammation or edema.
  2. $FA_{FWcorr}$ (FW-corrected Fractional Anisotropy): Representing the actual structural integrity of the white matter tissue.

Methodology and Data Harmonization

The study leveraged an unprecedented dataset of 9,671 imaging sessions from 4,702 participants across nine longitudinal cohorts, including ADNI, BLSA, BIOCARD, NACC, ROS/MAP/MARS, VMAP, and WRAP.

To ensure consistency across different sites and MRI scanners, the team utilized Longitudinal ComBat. This statistical technique removes "batch effects" (technical variations between machines) while preserving the underlying biological trends over time. The researchers then mapped these metrics across 48 specific white matter tracts, categorized into association, limbic, projection, and transcallosal pathways.

Key Findings: Distinct Biological Signatures

The most significant discovery was that different risk factors appear to "attack" white matter through different biological pathways.

Sex and Racial Disparities

Sex and race were primarily associated with changes in tissue integrity ($FA_{FWcorr}$).

  • Sex: Females generally exhibited lower tissue integrity across 33 tracts compared to males, with the most pronounced differences found in projection tracts (e.g., ventral premotor projection).
  • Race: Non-Hispanic Black adults generally showed lower tissue integrity in 32 tracts compared to White adults, particularly in the superior longitudinal fasciculus.

The Genetic Influence of APOE-ε4

In contrast, the APOE-ε4 genotype—the strongest genetic risk factor for late-onset AD—was primarily associated with Free-Water (FW). Carriers of the allele showed significantly higher fluid levels in limbic and occipital tracts, such as the ILF and Cingulum. Interestingly, the gene had minimal impact on the actual tissue integrity metric ($FA_{FWcorr}$), suggesting that genetic risk may drive fluid-related changes like neuroinflammation rather than direct structural degradation.

Regional Exceptions and Intersectionality

The study highlighted that brain aging is not uniform. While White adults generally had better metrics across most regions, the fornix—a key memory circuit—presented an exception. In this specific area, White adults showed higher free-water and lower tissue integrity than Black adults.

Furthermore, the research demonstrated that these factors do not exist in isolation. Race was found to modify the association between sex and tissue integrity in certain projection fibers, emphasizing that the biological experience of aging is shaped by complex interactions between multiple risk factors.

Limitations and Future Directions

The authors noted a "healthy volunteer" bias: Black participants in these cohorts were statistically less likely to be cognitively impaired than their White counterparts, which is the inverse of general population trends. This suggests that the racial disparities observed in this study may actually be an underestimation of the real-world gap.

While this research identifies what differences exist, the next step is determining why. The authors suggest that future investigations must look beyond innate biology to social determinants of health—including socioeconomic status, education, and the chronic stress of discrimination—to fully understand the drivers of these disparities in brain health.

Summary updated

Abstract

Sex and apolipoprotein E (APOE) ε4 carrier status relate to white matter microstructural integrity. Females generally have lower free water-corrected fractional anisotropy compared to males. APOE ε4 carriers tended to have higher free water than non-carriers.
BibTeX
@inproceedings{peterson2024sex,
  author = {Amalia Peterson and Aditi Sathe and Dimitrios Zaras and Yisu Yang and Alaina Durant and Kacie Deters and Niranjana Shashikumar and Kimberly R. Pechman and Michael E. Kim and Chenyu Gao and Nazirah Mohd Khairi and Zhiyuan Li and Tianyuan Yao and Yuankai Huo and Logan Dumitrescu and Katherine A. Gifford and Jo Ellen Wilson and Francis E. Cambronero and Shannon L. Risacher and Lori L. Beason‐Held and Yang An and Konstantinos Arfanakis and Guray Erus and Christos Davatzikos and Duygu Tosun and Arthur W. Toga and Paul M. Thompson and Elizabeth C. Mormino and Mohamad Habes and Di Wang and Panpan Zhang and Kurt G. Schilling and Marilyn Albert and Walter A. Kukull and Sarah Biber and Bennett A. Landman and Sterling C. Johnson and Julie A. Schneider and Lisa L. Barnes and David A. Bennett and Angela L. Jefferson and Susan M. Resnick and Andrew J. Saykin and Timothy J. Hohman and Derek B. Archer and Andrew J. Saykin and Timothy J. Hohman and Derek B. Archer},
  title = {{Sex and <i>APOE</i> ε4 allele differences in longitudinal white matter microstructure in multiple cohorts of aging and Alzheimer's disease}},
  booktitle = {Alzheimer},
  year = {2024},
  doi = {10.1002/alz.14343},
}
Figure