White Matter Microstructural Changes Linked to Declining Cognitive Performance, Memory Issues

Derek B. Archer, PhD

A large-scale study spanning nearly 2 decades across 9 cohorts provided further evidence that white matter microstructural changes, particularly free water (FW), are associated with cognitive performance and decline, especially memory. These data raise the importance of FW correction in dMRI studies and highlight the limbic system, especially the cingulum and fornix, as key regions associated with cognitive decline.¹

Published in JAMA Neurology, the trial featured data on 4467 participants (2698 female [60.4%]) aged 50 years and older who underwent 9208 cognitive sessions between September 2002 and November 2022. The cohort featured mostly cognitively unimpaired (n = 3213) patients, as well as those with mild cognitive impairment (MCI; n = 972), and few with Alzheimer disease (AD) dementia (n = 282).

Among the main findings was the fact that higher global FW was associated with lower performance across all cognitive domains. For memory, executive function, and language, which all showed global associations, the top links were related to fornix FW (ß = –1.069; P <.001), TC angular gyrus (ß = –0.512; P <.001), and cingulum FW (ß = –0.776; P <.001), respectively. To the authors knowledge, this was the first large-scale study to evaluate the association between FW-corrected white matter metrics and cognitive performance and decline using a large-scale harmonized dataset.

"Our cross-sectional analysis found that FW exhibited global associations with cognitive performance along the cognitive impairment spectrum, consistent with previous findings of widespread white matter degeneration in cognitive aging," senior author Derek B. Archer, PhD, an assistant professor of neurology at Vanderbilt University Medical Center, and others, wrote.

They continued on to note that "while higher FW was consistently associated with lower performance across all cognitive domains, intracellular metrics showed limited associations with memory but were broadly significant for executive function and language, excluding sensorimotor projection tracts. This suggests that FW, potentially reflecting neuroinflammation or atrophy, may serve as a broader indicator of cognitive decline, whereas intracellular changes, potentially reflecting overt tissue abnormalities, appear to play more domain-specific roles, particularly in executive function and language."

READ MORE: Phase 2 GABriella Study Represents First Step in Testing Selective Novel Y-Secretase Modulators for Prodromal Alzheimer Disease

In a bootstrapped analysis, the covariate-only model explained 34.93% (95% CI, 32.06–37.79) of variance in memory, 29.30% (95% CI, 26.68–32.18) in executive function, and 27.76% (95% CI, 25.05–30.27) in language. Among microstructural metrics, free water (FW) accounted for the largest additional variance beyond covariates for memory (ΔR² = 1.55; 95% CI, 0.04–3.27), executive function (ΔR² = 1.52; 95% CI, 0.14–2.95), and language (ΔR² = 1.90; 95% CI, 0.51–3.42) decline.

In a tract-type analysis categorizing FW into projection, motor TC, prefrontal TC, association, occipital TC, parietal TC, and limbic tracts, limbic tracts contributed the greatest additional variance beyond covariate-only models for decline in memory (ΔR² = 2.07; 95% CI, 0.41–4.04), executive function (ΔR² = 1.71; 95% CI, 0.31–3.17), and language (ΔR² = 2.37; 95% CI, 0.91–3.86).

"Further evaluation of FW by tract type highlighted the limbic tracts as being most predictive of subsequent cognitive decline, consistent with the established role of limbic structures in memory and emotion regulation," Archer et al wrote in the discussion section.¹ "Specifically, the cingulum and fornix were strongly associated with memory decline, echoing findings that these tracts are vulnerable in aging. Overall, our findings suggest that microstructure in the fornix and cingulum may be valuable biomarkers for aging and AD, with targeted focus on these regions offering potential for more effective interventions to support healthy aging."

Investigators also found synergistic associations between white matter abnormalities and other AD endophenotypes that relate to cognitive performance and decline. Among 2826 cross-sectional interaction tests, the strongest effects were observed for fornix FW with hippocampal volume (ß = 10.598; P <.001), cingulum bundle FW with SPARE-AD index (ß = –0.532; P <.001), and inferior temporal gyrus TC tract FW with baseline diagnosis (ß = –0.537; P <.001), each predicting poorer memory performance.

REFERENCE
1. Peter C, Sathe A, Shashikumar N, et al. White Matter Abnormalities and Cognition in Aging and Alzheimer Disease. JAMA Neurol. 2025;82(8):825-836. doi:10.1001/jamaneurol.2025.1601