The sacral cord conduction time of the soleus H-reflex was investigated in 30 normal adult subjects using three different methods. (1) The posterior tibial nerve was stimulated at the popliteal fossa by graded electric shocks, and the recordings were made from different lumbar epidural intervertebral levels. The afferent action potentials from the dorsal roots and the reflexively evoked efferent action potentials from the ventral roots were recorded. The time interval between the negative peaks of the ventral and dorsal root potentials was used to calculate the approximate sacral cord reflex delay time, which was found to be 1.3 ms. on average. (2) The sacral cord reflex delay time was found to be about 2.0 ms using the conduction time of the afferent, that of the efferent limbs and total reflex time of the soleus H-response. (3) By stimulating the lumbosacral roots at the epidural levels and using the difference between the soleus H and M response latencies, the sacral cord reflex delay was determined to be approximately 2.4 ms. These findings indicated that the soleus H-reflex is exclusively monosynaptic. It is proposed that in humans the synaptic transmission at the sacral cord is approximately 0.4 ms.

Segmental spinal cord and root potentials in response to stimulations of the dorsal nerve of penis, tibial nerve, and the sural nerve were epidurally recorded in normal subjects. EMG responses from the bulbocavernosus (BC) and the various leg muscles were also recorded in response to bipolar stimulations by the same epidural needle electrodes of the sacral cord and lumbosacral roots. The afferent conduction velocity from the penis to Th12-L1 intervertebral level was about 40 m/sec on the average, which is significantly slower than those obtained by the stimulation of the mixed and cutaneous nerves at the lower limb. The latency of the motor responses of the BC muscle from Th12-L1 spine levels were found comparatively longer than those of thigh muscles on maximal epidural stimulation in spite of the shorter distances to the BC muscle. The central conduction delay within the sacral cord of the bulbocavernosus reflex was calculated and found to be about 8.2 msec, while the central conduction time was about 1.1 msec for the Soleus-H-Reflex. These findings may suggest that there may be about 5-6 synapsis necessary for the first component of the bulbocavernosus reflex, though some faster oligosynaptic cord linkage may also exist.

Needle electrical stimulation of the lumbosacral roots at the laminar level of the Th12‐L1 or L1‐2 intervertebral spaces were performed in 24 normal subjects and 58 patients with various kinds of lumbar radiculopathy (unilateral L4, L5 and S1 herniated nucleus pulposus and lumber stenosis). The root stimulation method was compared with conventional needle EMG. Lumber electrical stimulation showed root abnormalities objectively in 80% of patients while the diagnostic value of needle EMG was 65%. Therefore, electrical root stimulation is superior to routine EMG for localizing lumbar root involvement. However, the only needle EMG demonstrated the root pathology in 7 cases (12%) and single electrophysiological abnormality was found by the root stimulation in 16 cases (27%). Thus, both electrophysiological methods should be complementary to each other in evaluation of the lumbar radiculopathy.