Time to reconfigure balancing behaviour in man: changing visual condition while riding a continuously moving platform.

While balancing on a continuously anteroposterior (A-P) translating platform (10 cm, 0.5 Hz), the head normally oscillates with the platform without vision but is stabilized in space with vision. We estimated the time to shift from one to the other balancing behaviour when visual condition changed at some stage during the balancing trials. Ten subjects performed randomly 50 balancing trials (each lasting 18 s): 10 trials with eyes open (EO), 10 with eyes closed (EC), 15 in which participants started with EO and closed their eyes (condition EO!EC) in response to an acoustic signal delivered during the trial, and 15 starting with EC and closing their eyes (EC!EO) in response to the same signal. No other speciWc instruction was given. Displacements of malleolus, hip and head, and EMG from leg and axial muscles were recorded. Indexes of amplitude of A-P head and hip oscillation and of amplitude of EMG activity were computed. All variables were larger with EC than EO. On changing visual condition during the trial, the pattern of head and hip movement and of muscle activity turned into that appropriate for the new visual condition in a time-interval ranging from about 1 to 2.5 s. For each subject, the mean latency of the change in the balancing behaviour was assessed by statistical methods. On average, the latencies of kinematics and EMG changes proved to be longer for the EO!EC condition than viceversa. Further, the latencies of the changes were also measured across all EO!EC and EC!EO individual trials. These values were clustered around particular epochs of the Wrst few oscillation cycles following the shift in visual condition. The results show that subjects can rapidly adapt their balancing behaviour to the new visual condition. However, they appear to refrain from releasing the new behaviour were this unWt, and unfastened it at appropriate time in the next platform translation cycle. These Wndings reveal the temporal and spatial features of the automatic release of the new balancing strategy in response to a shift in the ongoing sensory set, and emphasize the swiftness in the change in balancing behaviour when subjects pass from a non-visual to a visual reference frame.