Microstructural Brain Reserve Moderates the Associations Between 24-h Movement Behaviors and Cognition in Older Adults: Cross-Sectional Findings From the AGUEDA Trial.

No direct associations were observed between 24-hour movement behaviors and any cognitive domain in cognitively unimpaired older adults.

• Sample included 89 cognitively unimpaired older adults (mean age 71.61 ± 3.85 years; 56.2% females) from the AGUEDA trial (NCT05186090).
• Movement behaviors assessed included MVPA, LPA, sedentary behavior, and sleep, measured via 9-day wrist-worn accelerometry.
• Cognitive function was assessed across five domains: attention/inhibitory control, episodic memory, processing speed, visuospatial processing, and working memory.
• All p-values for direct associations were > 0.05.
• Compositional linear regression models (CoDA) were used, adjusting for age, sex, and education.

Gray matter mean diffusivity (GMMD) signature, but not thickness/volume signature, moderated several movement behavior-cognition associations.

• Two Alzheimer's disease brain signatures were derived: GMMD signature (microstructural integrity) and thickness/volume signature (macrostructural integrity) from cortical and hippocampal regions.
• GMMD moderated the associations between movement behaviors and cognition across multiple domains.
• The thickness/volume signature did not significantly moderate any movement behavior-cognition associations.
• This differential moderation effect suggests that microstructural, rather than macrostructural, brain reserve is the relevant moderator.

Moderation by GMMD was most evident for episodic memory, with significant interactions for MVPA, LPA, and sleep.

• All p-values for interaction terms involving episodic memory were ≤ 0.04.
• MVPA, LPA, and sleep each showed significant interactions with GMMD signature in predicting episodic memory.
• Episodic memory showed the broadest range of movement behaviors that interacted with GMMD compared to other cognitive domains.

Light physical activity (LPA) exhibited the broadest pattern of moderation by GMMD across cognitive domains.

• LPA interacted significantly with GMMD signature across four cognitive domains: attention/inhibitory control, episodic memory, processing speed, and working memory.
• All p-values for LPA interaction terms were ≤ 0.04.
• LPA showed a wider range of moderated cognitive associations than MVPA, sedentary behavior, or sleep.

The study used a compositional data analysis (CoDA) framework to examine movement behavior-cognition associations, treating 24-hour behaviors as an integrated composition.

• CoDA compositional linear regression models were applied to respect the co-dependent nature of 24-hour time-use behaviors (MVPA, LPA, SB, and sleep summing to 24 hours).
• Movement behaviors were objectively assessed over 9 days using wrist-worn accelerometry.
• Models adjusted for age, sex, and education as covariates.
• Interactions between movement behavior compositions and each brain signature were formally tested.

The findings support a personalized approach to promoting cognitive health in older adults based on microstructural brain vulnerability profiles.

• Results suggest that the cognitive benefits of movement behavior changes may differ depending on an individual's microstructural brain integrity.
• The authors call for future prospective longitudinal studies and CoDA interventional trials to establish temporal precedence.
• The authors suggest evaluating whether targeted time reallocations can attenuate cognitive trajectories across distinct microstructural vulnerability profiles.
• The cross-sectional design of the study precludes causal inference.