Joint trajectories of brain atrophy, white matter hyperintensities and cognition quantify brain maintenance.

Quantifying the Preservation of the Aging Brain

Researchers report that changes in brain atrophy and white matter hyperintensities additively and independently predict cognitive change. This finding addresses a central mystery in neurocognitive aging: "brain maintenance." This term refers to the preservation of brain structure and function relevant to cognitive performance. While we know brain shrinkage and vascular damage correlate with cognitive loss, quantifying how much cognitive health stems directly from structural integrity has remained difficult.

A new study from the DELCODE cohort proposes a solution. By modeling how brain shrinkage, white matter changes, and thinking skills evolve together over four years, the authors created a new index. This index shows how much cognitive health is protected by structural brain integrity. They found that factors like personality and lifestyle correlate with this maintenance, providing a measurable way to track how individuals hold onto their cognitive resources.

Defining Brain Maintenance

The core challenge in neurocognitive aging is that the brain does not age uniformly. Some individuals possess an inherent or acquired ability to attenuate the changes typically associated with aging. To quantify this, the authors move away from looking at single markers in isolation. Instead, they treat the brain as a coupled system.

Rather than simply asking how much the brain has shrunk, the researchers ask to what extent cognition is preserved because the brain remains structurally intact. To do this, they focus on two primary structural markers. The first is the medial temporal lobe to ventricle ratio (MTLV-ratio). This tracks aging-related atrophy (the wasting away of tissue). It is calculated by dividing the volume of the medial temporal lobe (a region critical for memory) by the sum of the MTL volume and the volume of the inferior lateral ventricles (fluid-filled spaces that expand as tissue shrinks).

The second marker is white matter hyperintensities (WMH). These appear as bright spots on MRI scans. They serve as a proxy for cerebrovascular abnormalities (damage to the brain's wiring caused by small vessel disease).

Modeling the Co-evolution of Structure and Thought

To capture these moving parts, the authors utilized Trivariate Latent Growth Curve Modelling (LGCM). Think of LGCM as a way to fit multiple moving lines to a single dataset simultaneously. Rather than analyzing WMH, atrophy, and cognition as three separate problems, LGCM treats them as a unified, evolving system. This allows researchers to estimate the starting point (the intercept) and the rate of change (the slope) for each domain. It also shows how changes in one domain correlate with changes in another.

As illustrated in, the model tracks the relationship between WMH, MTLV-ratio, and the Preclinical Alzheimer’s Cognitive Composite (PACC5). The PACC5 is a score designed to detect early cognitive shifts. On average, the study finds that WMH volumes increase and MTLV-ratios decrease. While the PACC5 score showed only a weak, non-significant increase across the population, the authors note substantial inter-individual variability. This means that while the group average appeared stable, many individuals experienced significant changes.

The researchers' most critical finding involves the unique contributions of these structural changes. Through robust multiple linear regression, the authors demonstrate that changes in MTLV-ratio and WMH additively and independently predict cognitive change .

This means vascular damage (WMH) and structural atrophy (MTLV-ratio) represent distinct pathways of decline. Specifically, the study finds that the rate of MTLV-ratio decline and the progression of WMH both exert independent pressures on cognitive performance.

By combining these structural trajectories, the authors derived a "brain maintenance index." This index represents the portion of an individual's cognitive trajectory that can be explicitly attributed to preserved brain structure.

Identifying the Drivers of Resilience

Once the index was established, the researchers looked for factors associated with differences in maintenance. They found that personality and lifestyle factors correlate with how well a person maintains their brain.

The study reports that certain personality traits are linked to unfavorable trajectories. For instance, higher levels of neuroticism (a tendency toward emotional instability) and increased depressive symptoms were associated with steeper MTLV-ratio decline and poorer brain maintenance .

Conversely, higher "openness"—a trait associated with curiosity and engagement—was linked to better maintenance.

The researchers also found correlations between lifestyle choices and this index. Poor sleep quality was associated with lower cognitive performance changes. The authors suggest that mental health and cerebrovascular management are central to this equation. Crucially, the study validates the brain maintenance index by showing it correlates with DunedinPACE (a molecular marker of biological aging pace) .

This connection suggests the index reflects a person's overall biological aging process.

Limits of the Framework

Despite the strength of the model, the authors identify several boundaries. Because the LGCM framework focuses on co-evolution, it cannot determine temporal sequencing. It cannot definitively say if WMH causes atrophy or if they simply evolve in tandem. To resolve this "lead-lag" relationship, future studies would require different modeling approaches.

Furthermore, the study's cohort had relatively low vascular risk. The authors note this might lead to an underestimation of associations involving WMH. The reliance on self-reported questionnaires for lifestyle and personality also introduces potential bias. Finally, the four-year follow-up period may be too short to capture the long-term impact of midlife lifestyle interventions on brain reserve.

Figures from the paper