This study compared the attentional demands between cane-free walking and cane walking in patients with stroke during the transitional period of cane weaning. Patients with stroke who had just learned to walk cane-free were recruited. Cross-sectional measurement was scheduled within 30 days since the patients were able to walk independently without a quad cane. The dual-tasking paradigm required participants to walk with and without a cane, as well as perform continuous subtractions by 1 s (low-demand) or 3 s (high-demand). The cognitive-motor interference (CMI) of walking velocity was calculated as [(low-demand - high-demand)/low-demand] x 100%. Nine participants (average age, 53.4 +/- 6.4 years; stroke onset, 38-131 days) were recruited, and eight showed positive CMI. The paired t-test confirmed a significantly smaller CMI during cane-free walking than during cane walking [t (8) = -3.168; P = 0.013]. The Pearson correlation tests revealed associations between age and CMI of cane walking (r = 0.751; P = 0.010) and CMI of cane-free walking (r = 0.584; P = 0.050). The time since independent cane-free walking was associated with CMI of cane walking (r = 0.699; P = 0.018). In conclusion, experience with cane-free walking leads to increased attentional demand for cane walking. In subacute stroke patients weaning use of a cane, the attentional demand for cane-free walking decreases to less than that of cane walking. During both cane and cane-free walking, the older the participant, the more the walking performance deteriorated due to dual-tasking.
