We performed transcranial magnetic stimulation (TMS) to elucidate the D- an
d I-wave components comprising the motor evoked potentials (MEPs) elicited
from the leg motor area, especially at near-threshold intensity. Recordings
were made from the tibialis anterior muscle using needle electrodes. A fig
ure-of-eight coil was placed so as to induce current in the brain in eight
different directions, starting from the posterior-to-anterior direction and
rotating it in 45 degrees steps. The latencies were compared with those ev
oked by transcranial electrical stimulation (TES) and TMS using a double co
ne coil. Although the latencies of MEPs ranged from D to I3 waves, the most
prominent component evoked by TMS at near-threshold intensity represented
the I1 wave. With the double cone coil, the elicited peaks always represent
ed I1 waves, and D waves were evoked only at very high stimulus intensities
, suggesting a high effectiveness of this coil in inducing I1 waves. Using
the figure-of-eight coil, current flowing anteriorly or toward the hemisphe
re contralateral to the recorded muscle was more effective in eliciting lar
ge responses than current flowing posteriorly or toward the ipsilateral hem
isphere. The effective directions induced I1 waves with the lowest threshol
d, whereas the less effective directions elicited I1 and I2 waves with a si
milar frequency. Higher stimulus intensities resulted in concomitant activa
tion of D through I3 waves with increasing amount of D waves, but still the
predominance of I1 waves was apparent. The amount of I waves, especially o
f I1 waves, was greater than predicted by the hypothesis that TMS over the
leg motor area activates the output cells directly, but rather suggests pre
dominant transsynaptic activation. The results accord with those of recent
human epidural recordings. (C) 2000 Elsevier Science B.V. All rights reserv
ed.