Postural electromyographic responses in the arm and leg following galvanic vestibular stimulation in man

Application of a small (around 1 mA), constant electric current between the mastoid processes (galvanic stimulation) of a standing subject produces enhanced body sway in the approximate direction of the ear behind which the anode is placed. We examined the electromyographic (EMG) responses evoked by such stimulation in the soleus and in the triceps brachii muscles. For soleus, subjects stood erect, with their eyes closed, leaning slightly forward. The head was turned approximately 90° to the right or left relative to the feet. In averaged records (n=40), current pulses of 25 ms or longer modulated the EMG in a biphasic manner: a small early component (latency 62±2.4 ms, mean ± SEM) was followed by a larger late component (latency 115±5.2ms) of opposite sign, which was appropriate to produce the observed body sway. The early component produced no measurable body movement. Lengthening the duration of the stimulus pulse from 25 to 400 ms prolonged the late component of the response but had little effect on the early component. Short- and long-latency EMG responses were also evoked in the triceps brachii muscle if subjects stood on a transversely pivoted platform and had to use the muscle to maintain their balance in the anteroposterior plane by holding a fixed handle placed by the side of their hip. The latency of the early component was 41±2.6 ms; the latency of the late component was 138±4.3 ms and was again of appropriate sign for producing the observed body sway. Galvanic stimulation evoked no comparable responses in either triceps brachii or soleus muscles if these muscles were not being used posturally. The responses were most prominent if vestibular input provided the dominant source of information about postural stability, and were much smaller if subjects lightly touched a fixed support or opened their eyes. The difference in latency between the onset of the early component of the response in arm and leg muscles suggests that this part of the response uses a descending pathway which conducts impulses down the spinal cord with a velocity comparable with that of the fast conducting component of the corticospinal tract. The late component of the EMG response occurs earlier in the leg than the arm. We suggest that it forms part of a patterned, functional response which is computed independently of the early component.