Real-world walking patterns are associated with regional brain atrophy and white matter lesions in middle-aged and older people: a Watch Walk-UK Biobank study

Clinic-based gait performance is associated with cognitive function and brain structure. This study investigated associations between multiple remotely collected digital gait biomarkers and brain structural changes in a large population-based cohort. Cross-sectional analysis of 6412 participants from the UK Biobank, with both valid wrist-worn accelerometer and brain MRI data. Digital gait parameters, including step count, usual and maximal gait speed, cadence, and stride regularity, were derived from accelerometry data using the validated Watch Walk algorithm. Brain volumes (total brain volume, white matter lesion volume, and frontal lobe, hippocampal, insular, and amygdala gray matter volumes) were measured using MRI. Multiple linear regression models were used to assess associations between gait parameters and brain volumes, adjusting for potential confounders, including demographic, lifestyle, and health-related factors. Restricted cubic splines explored the shape of associations. Faster maximal and usual gait speed and higher stride regularity were associated with lower white matter lesion volume and larger frontal lobe, hippocampal, insular, and amygdala gray matter volumes. Restricted cubic spline analyses revealed that the relationships between maximal gait speed and stride regularity, on one hand, and white matter lesion volume, frontal lobe, hippocampal, insular, and amygdala gray matter volumes, on the other hand, were predominantly linear. Gait speed and stride regularity, objectively measured using wrist-worn accelerometers in real-world settings, are significantly associated with brain structural integrity. These findings highlight the potential of digital gait biomarkers as scalable and cost-effective tools for monitoring brain health and identifying individuals at risk for cognitive decline.