Participants completed a supervised six-week programme comprising three 30-minute sessions per week (18 sessions; planned dose, 540 minutes). During each session, they synchronized their steps with a metronome while performing tandem walking with progressively demanding object-carrying tasks. Participants maintained their usual walking speed and were instructed to give equal attention to gait and the concurrent motor task. Progression was designed by the research team in accordance with general motor-learning principles and involved predefined weekly increases in metronome tempo and concurrent motor-task complexity.

The intervention combined RAS with progressively intensified motor dual-task gait training.6 Participants completed a supervised six-week programme comprising three 30-minute sessions per week (18 sessions; planned dose, 540 minutes). During each session, they synchronized their steps with a metronome while performing tandem walking with progressively demanding object-carrying tasks. Participants maintained their usual walking speed and were instructed to give equal attention to gait and the concurrent motor task. Progression was designed by the research team in accordance with general motor-learning principles and involved predefined weekly increases in metronome tempo and concurrent motor-task complexity. provided that the preceding stage had been completed safely. All sessions were delivered individually by the same licensed physiotherapist experienced in neurology.

Participants completed a 10-m walk at their comfortable speed under single-task and motor dual-task conditions. In the dual-task condition, they carried a ball while walking and were instructed to give equal attention to both tasks. Three trials were performed under each condition, and mean completion time was used in the analysis. Positive values indicated greater dual-task interference.

Functional mobility was assessed using the Timed Up and Go test, with shorter completion times indicating better performance. One familiarization trial preceded the timed assessment.

Dynamic balance was assessed using the 14-item Mini-BESTest, which evaluates anticipatory postural adjustments, reactive postural control, sensory orientation, and dynamic gait. Total scores range from 0 to 28, with higher scores indicating better balance. Static balance was assessed bilaterally using the Single-Leg Stance Test. Two trials were performed on each side, and the longest duration before foot contact, stance-foot movement, or loss of balance was recorded.