INTENTIONAL ONLINE CONTROL OF PROPULSIVE FORCES IN HUMAN GAIT

In locomotion, the capability to control and modulate intentionally th e propulsive forces is fundamental for the adaptation of the body's pr ogression, both in speed and direction. The purpose of this experiment was to determine how human beings can achieve such control on-line. T o answer this question, four subjects walking steadily were faced with a linear increase in resistance (impeding forward displacement), last ing 3 s, once per minute. At the end of the variation, the new resista nce was maintained. There were two tasks; in both tasks, in the initia l steady state, the subjects had to walk steadily at 1.3 m s(-1). As t he resistance increased, subjects were either required to maintain the ir walking speed (compensation task) or to let the walking speed and a mplitude adapt freely (no-intervention task). This provided an estimat e of the effects of the perturbation alone. Throughout the experiment, the stride frequency (114 step min(-1)) was fixed by a metronome. Sub jects maintained their stride frequency on both tasks. In the no-inter vention task, walking speed was 1.3 and 1 m s(-1) under normal and hig h resistance respectively. In the compensation task, under high steady resistance, walking speed was maintained by an increase in the activa tion gain of the neuromuscular synergy: all recorded muscles increased their EMG activity, but without any change in the shape of their acti vation profile throughout the cycle. During the transitional phases, h owever, as the resistance began to increase, the walking speed decreas ed temporarily (-2%) before returning rapidly to its initial value. By contrast, at the end of the resistance increase, no such changes in s peed were observed. During the transitional phases, the on-line compen sation for the resistance increase induced modifications in the shape of the activation burst in the medial gastrocnemius such that the tran sitional cycles clearly differed from the steady state cycles. The res ults observed in the compensation task suggest that the subjects used two different modes of control during steady states and transitional p hases. In stable dynamic conditions, there appears to be an ''intermit tent control'' mode, where propulsive forces are globally managed for the entire stance phase. As a result, no compensation occurred at the beginning of the perturbation. During the resistance increase, subject s appeared to switch to an ''on-line control'' mode in order to contin uously adapt the propulsive forces to the time course of the external force, resulting in an observable compensation at the end of the resis tance change.