Attentional requirements of walking according to the gait phase and onset of auditory stimuli

A dual-task paradigm was used to examine the influence of an attention demanding cognitive task on each phase of gait. Twenty-three participants (aged 18–27) walked on a treadmill at a 20% increase of their self-selected speed, either alone or while performing a cognitive task. Muscle activity was measured with electromyography (iEMG) for eight muscles of the dominant leg. The cognitive task consisted of subtracting one (EASY) or seven (HARD) from aurally presented numbers. Reaction time (RT) and accuracy were recorded. iEMG events were selected according to stimulus onset (0–150 ms, 150–300 ms and 300–450 ms) prior to phases of gait (double-leg stance, single-leg stance and swing). There was a decrease in iEMG amplitude of fibularis longus (p = .013) and a trend in the same direction for vastus lateralis (p = .065) while walking and performing the cognitive task. When stimulus onset was considered, iEMG of medial gastrocnemius (p = .021) and lateral gastrocnemius (p = .004) were reduced during single-leg stance, when stimuli occurred between 300 and 450 ms prior to this phase. Cognitive performance was affected by task difficulty (RT, accuracy) and by dual-task load (RT). Dual-task costs were observed in both the motor and the cognitive tasks, suggesting that walking requires attention. There was a specific moment (300 ms after stimulus onset) during single-leg stance when dual-task costs were most pronounced, corroborating supraspinal involvement in the control of normal walking. Time-based approaches should be considered when analyzing attentional demands of a dynamic task such as gait.