Associations of Gait Speed with Neuroimaging Biomarkers: Findings from the LifeAfter90 (LA90) Study

TL;DR

This study explores how walking speed relates to brain changes in very old adults, finding that faster walking may indicate better brain health in women.

Contribution

The study reveals sex-specific associations between gait speed and white matter integrity in the oldest old, a rarely studied population.

Findings

Faster gait speed was associated with lower white matter hyperintensities in females but not in males.

Gait speed showed non-significant trends with amyloid accumulation and white matter integrity in the overall cohort.

No significant associations were found between gait speed and hippocampal volumes.

Abstract

Gait speed is a well‐known marker of cognitive function and mortality in older adults. However, its association with neuroimaging biomarkers is understudied. We tested associations between gait speed and biomarkers of neurodegeneration, cerebrovascular injury and amyloid accumulation in the oldest old.

The LifeAfter90 Study (2018‐2022) included a neuroimaging subsample of 103 participants aged 90 and older. Gait speed (meters/second) was measured twice per visit using the 4‐Meter Walk Test and we used the average gait speed score from the visit closest to each participant's neuroimaging scan date. Hippocampal volumes (total, right and left hippocampus) and measures of white matter integrity (log‐white matter hyperintensities (WMH), fractional anisotropy (FA)) were measured with 3T MRI and amyloid standardized uptake value ratio (SUVR) with florbetapir PET. Hippocampal volumes and WMH were adjusted for intracranial volume. Linear regression models tested associations between gait speed and brain imaging markers, adjusting for age, sex/gender, race/ethnicity, and education. Sex‐stratified models were used to explore potential differences in these associations by sex.

Mean age of participants was 92.5(±2.3) years, and 59.2% were women (Table 1). Overall, gait speed was not significantly associated with any brain regions of interest (Figure 1). Although not statistically significant, faster gait speed suggested better white matter integrity (βlogWMH = ‐0.59, 95% CI ‐1.28, 0.10; βFA =0.27, 95% CI ‐0.54, 1.09) and lower amyloid‐SUVR (βamyloid= ‐0.10, 95% CI ‐0.33, 0.12). In contrast, it was associated with lower hippocampal volumes (βtotal = ‐0.30, 95% CI ‐1.24, 0.64; βright = ‐0.13, 95% CI ‐1.04, 0.79; βleft = ‐0.41, 95% CI ‐1.37, 0.55) (Figure 1). In the stratified analysis, faster gait speed was significantly associated with lower WMH volume in females (βlogWMH = ‐1.00, 95% CI ‐1.91, ‐0.10) but not in males (βlogWMH = ‐0.12, 95% CI ‐1.46, 1.21).

Although gait speed was not associated with specific neuroimaging biomarkers in the overall cohort, faster gait speed was associated with lower WMH in females. This study underscores the need to further investigate gait speed as an accessible indicator of neuropathological changes, especially in 90+ populations where neuroimaging might be challenging.

Figures