[Skip to Navigation]
October 1978

Central Latencies of Somatosensory Cerebral Evoked Potentials

Author Affiliations

From the Division of Clinical Electrophysiology, Department of Neurology, University of Iowa College of Medicine, Iowa City (Drs Kimura and Yamada); and the Department of Neurosurgery, Neurological Institute, Tokyo Women's Medical College, Tokyo (Dr Kawamura).

Arch Neurol. 1978;35(10):683-688. doi:10.1001/archneur.1978.00500340059012

• Latencies of the nuchal and scalp-recorded somatosensory evoked potential (SEP) were determined by stimulation of the median nerve at the wrist and peroneal nerves at the knee in 27 healthy subjects. With stimulation of the median nerve, the estimated conduction time from the wrist to C7 (11.4 ± 1.1 ms) was 2.1 1.1 ms less than the latency of the major negative peak, N0, of the nuchal SEP (13.6 ± 1.0 ms). The latter potential, in turn was 0.9 ± 0.6 ms less in latency compared with the initial positive peak, P0, of the scalp-recorded SEP (14.2 ± 1.0 ms). Even if only a minimal delay of 1.5 ms at the three synapses was allowed, the calculated central latency measured to P0 (2.9 ± 0.9 ms) was too small; however, that measured to N1 (6.9 ± 1.9 ms) was appropriate to cover the distance from C7 to the scalp (24.6 ± 1.9 cm).

These findings suggest generation of N0 primarily from the dorsal column nuclei and P0 from either lemniscal inflow or cerebellar or thalamic nuclei. It is thus likely that N1 is the initial cortical discharge of the somatosensory impulse. Latency determination was more reliable with the median-evoked than the peroneal-evoked SEP. Of various latency computations, right-left comparison in the same subject and difference between successive peaks among different subjects were least variable. These measures, therefore, may best be used to assess neural conduction in the somatosensory pathways.

Add or change institution