For stroke survivors navigating daily life, the simple act of walking while thinking — checking a phone, holding a conversation, scanning for traffic — is far more cognitively taxing than for healthy peers. Understanding precisely which mental demands most disrupt walking could reshape how rehabilitation programs are designed and prioritized, shifting focus from generic dual-task training toward targeted cognitive-motor interventions.
This cross-sectional laboratory study enrolled 53 individuals with chronic stroke and had them complete the Stand and Walk for a 20-meter Round Trip test under seven distinct dual-task conditions spanning executive function, sustained attention, working memory, verbal fluency, and visuospatial processing. Gait parameters were captured via wearable motion sensors while cognitive output was simultaneously quantified as correct responses per second. Dual-task costs were then calculated for both gait speed and cognitive performance across all conditions. Significant gait speed reductions emerged universally across all seven dual-task conditions compared to single-task walking, but the magnitude of interference was not uniform — working memory-demanding tasks produced the steepest gait decrements, suggesting a domain-specific hierarchy of cognitive-motor interference in this population.
These findings build meaningfully on prior dual-task research in stroke, which has largely treated cognitive load as a monolithic variable. The domain-specificity observed here aligns with neuroimaging evidence that prefrontal-cerebellar circuits governing working memory substantially overlap with locomotor control networks — a convergence particularly vulnerable after cerebrovascular injury. From a rehabilitation standpoint, identifying which cognitive domains impose the greatest walking cost matters practically: programs targeting working memory under ambulatory conditions may yield disproportionate functional gains. Key limitations include the cross-sectional design, which precludes causal inference, a single-center laboratory setting that may not reflect community ambulation, and a sample of 53 that limits subgroup analyses by lesion location or severity. This is incremental but well-constructed science that usefully operationalizes cognitive-motor interference in a clinically meaningful framework.
