Sensorimotor integration during stance: processing time of active or passive addition or withdrawal of visual or haptic information.

Vision (V) and touch (T) help stabilize our standing body, but little is known on the time-interval necessary for the brain to process the sensory inflow (or its removal) and exploit the new information (or counteract its removal). We have estimated the latency of onset and the time-course of the changes in postural control mode following addition or withdrawal of sensory information and the effect of anticipation thereof. Ten subjects stood in tandem position. They wore LCD goggles that allowed or removed vision, or lightly touched (eyes-closed) with the index finger (haptic stimulation) a pad that could be suddenly lowered (passive task). In different sessions, sensory shifts were deliberately produced by opening (or closing) the eyes or touching the pad (or lifting the finger) (active task). We recorded eyelid movement and finger force (<1N), sway of center of foot pressure (CoP), electromyogram (EMG) of soleus, tibialis and peroneus muscle, bilaterally, and of extensor indicis. The latency of the CoP and EMG changes following the shifts were statistically estimated on the averaged traces of 50 repetitions per condition. Muscle activity and sway adaptively decreased in amplitude on adding stabilizing visual or haptic information. The time-interval from the sensory shift to decrease in EMG and sway was ∼0.5-2 s under both conditions. It was shorter for tibialis than peroneus or soleus and shorter for visual than haptic shift. CoP followed the tibialis by ∼0.2 s. Slightly shorter intervals were observed following active sensory shifts. Latencies of EMG and postural changes were the shortest on removal of both haptic and visual information. Subsequently, the time taken to reach the steady-state was ∼1-3 s under both active and passive tasks. A startle response at ∼100 ms could precede EMG changes. Reaction-time contractions in response to sensory shifts appeared at ∼200 ms, earlier than the adaptive changes. Changes in postural behavior require a finite amount of time from visual or haptic shift, much longer than reflexes or rapid voluntary responses, suggesting a time-consuming central integration process. This process is longer on addition than removal of haptic information, indicating a heavier computational load. These findings should be taken into account when considering problems of sensorimotor integration in elderly subjects or patients and when designing simulation models of human balance.