TL;DR
This study reveals that white matter hyperintensities (WMH) in the occipital lobe are linked to poorer long-term memory in older adults. This association appears to be driven by WMH causing atrophy in key medial temporal lobe structures, particularly the perirhinal cortex, which is crucial for memory.
What Problem Does This Paper Address?
White matter hyperintensities (WMH), visible on MRI, are indicators of small vessel cerebrovascular disease and are known to affect cognition in older adults. Simultaneously, atrophy in the medial temporal lobe (MTL), including the hippocampus, is a hallmark of aging and Alzheimer's disease, leading to memory decline. While both WMH and MTL atrophy are individually linked to cognitive issues, their specific interrelationship, particularly how WMH might contribute to MTL atrophy and subsequent memory loss, has been underexplored. This paper aims to bridge this gap by investigating the direct associations between regional WMH, MTL subregional volumes, and long-term memory performance in older adults without dementia.
Methods
The researchers employed a multi-pronged approach using MRI and neuropsychological testing.
- MRI Acquisition: High-resolution T1-weighted and T2-weighted FLAIR MRI scans were acquired from participants.
- WMH Segmentation: A deep learning-based approach using the ANTsX software ecosystem was utilized for segmenting WMH. This involved specialized ANTsPyNet functions for WMH segmentation and lobar parcellation based on the Desikan-Killiany-Tourville (DKT) atlas. The automated segmentation was manually refined by an expert rater for accuracy. WMH volumes were then square-root transformed to address skewness.
- MTL Subregional Volume Segmentation: The Automatic Segmentation of Hippocampal Subfields (ASHS) software was used to automatically segment MTL subregions, including hippocampal subfields (CA1, CA2, CA3, dentate gyrus, subiculum) and cortical MTL regions (entorhinal cortex, perirhinal cortex (BA35, BA36), and parahippocampal cortex). Volumes were adjusted for total intracranial volume (TIV).
- Neuropsychological Testing: The Rey Auditory Verbal Learning Test (RAVLT) was administered, with a focus on the delayed recall score as the primary outcome measure for long-term memory.
- Statistical Analysis: Linear regression models were used to assess associations between WMH and memory, WMH and MTL volumes, and MTL volumes and memory. Multiple comparisons were corrected using the Holm-Bonferroni method. Mediation analyses were conducted using the SPSS PROCESS macro to test if MTL subregional volumes mediated the relationship between WMH and delayed recall. All models were adjusted for age, sex, and education.
Datasets
The study utilized data from 139 older adults without dementia. These participants were recruited from three cohorts:
- The Biomarker Exploration in Aging, Cognition, and Neurodegeneration (BEACON) study.
- The UCI Alzheimer's Disease Research Center (ADRC) longitudinal cohort.
- The 90+ Study of the Oldest Old.
No new benchmarks were introduced; the study used existing datasets and established neuroimaging and neuropsychological assessment tools.
Key Insights and Ideas
- Regional Specificity Matters: The study highlights that not all WMH are equal in their impact. Only occipital WMH showed a significant association with memory decline, suggesting that the location of vascular damage is crucial.
- WMH as a Precursor to MTL Atrophy: The findings strongly suggest that WMH, particularly in the occipital lobe, can lead to structural changes (atrophy) in key memory-related brain regions within the MTL. This provides a potential mechanistic link between vascular health and neurodegeneration.
- Perirhinal Cortex as a Key Mediator: The perirhinal cortex (specifically Brodmann area 36) emerged as a critical mediator. This means that occipital WMH likely impair memory through their effect on reducing the volume of the perirhinal cortex. This underscores the importance of this region in the WMH-memory pathway.
- Beyond Global Measures: The study emphasizes the value of looking at regional WMH and subregional MTL volumes, rather than just global measures, to uncover more nuanced relationships.
Key Results
- Increased occipital white matter hyperintensities were significantly associated with worse Rey Auditory Verbal Learning Test (RAVLT) delayed recall performance (p = 0.040).
- Increased occipital WMH were also associated with reduced volumes in several MTL subregions: CA1 (p = 0.001), dentate gyrus (p < 0.001), perirhinal cortex (BA36) (p = 0.001), and parahippocampal cortex (PHC) (p = 0.007).
- Volumes of CA1, dentate gyrus, BA36, and PHC were all positively associated with better RAVLT delayed recall performance.
- A mediation analysis revealed that occipital WMH had an indirect effect on delayed recall, and this effect was fully mediated by the volume of the perirhinal cortex (BA36).
Limitations and Future Work
The authors acknowledge several limitations:
- Sample Demographics: The study sample was predominantly affluent, highly educated, and non-Hispanic white, limiting generalizability to more diverse populations.
- Cross-Sectional Design: The study's cross-sectional nature prevents the establishment of causal relationships. It can show associations but not definitively prove that WMH cause MTL atrophy and memory loss.
- Limited Clinical Status Examination: While participants were without dementia, the study had a limited number of individuals with mild cognitive impairment and no individuals with dementia. This restricted the ability to examine how clinical diagnosis might interact with the observed associations.
Future work suggested by the authors includes:
- Longitudinal Studies: Incorporating longitudinal data to track changes over time and better understand causal pathways.
- Diverse Samples: Recruiting more diverse populations to improve generalizability.
- Clinical Spectrum: Including a wider range of cognitive impairment statuses, including dementia, to explore these relationships across the clinical spectrum.
- AD Biomarkers: Integrating markers of Alzheimer's disease pathology, such as tau PET, to investigate the spatial colocalization of tau and MTL atrophy as a result of posterior WMH.
- Modifiable Risk Factors: Further examining modifiable vascular risk factors that could potentially reduce cerebrovascular burden and slow neurodegenerative processes.
Summary updated