Associations between structural network connectivity, gait and cognition in small vessel disease

Cerebral small vessel disease (SVD), including white matter hyperintensities (WMH) and lacunes of presumed vascular origin, affects gait by disrupting important white matter tracks. However, little is known about the relationship between the structural network connectivity and gait in subjects with cerebral SVD, and the interaction with cognitive function. We assessed gait characteristics (stride length, cadence and stride width, as well as Tinetti and Timed-Up-and-Go test) of 423 subjects with cerebral SVD. The structural network was constructed using diffusion tensor imaging and tractography. We applied graph-theory to calculate network efficiency from the undirected weighted network. Associations between network measures, conventional MRI markers for SVD, gait performances and cognitive index as global cognitive function were tested. Network measures were associated with conventional MRI markers for SVD (WMH and lacunes) and with gait performances. Stride length, Tinetti and Timed-Up-and-Go test were associated with network measures, independent of WMH or lacunes. However, after adjusting for cognitive index, these associations diminished and were not significant anymore. In the mediation analyses, cognitive index mediated the association between global efficiency and stride length, Tinetti and Timed-Up-and-go test. Regional analysis showed widespread involvement of cortical regions in gait performance, including frontal motor, cingulate and visuospatial regions, which diminished after adjusting for cognitive index. These results suggest that measures for network disruption were associated with gait performance, possibly indirectly by disrupting brain network responsible for cognitive function and hereby impairing gait performance.