Segmental demyelinating disease: reliability of nerve conduction measurements.

Serial nerve conduction studies were done in 5 patients with Guillain-Barre syndrome, 4 of whom were recovering and 1 who did not recover. In a total of 20 paired studies, the nerve conduction velocity deteriorated in spite of improvement in distal and proximal latencies in 5 cases, and nerve conduction velocity improved in spite of deterioration of latencies in 1 case. This seeming contradiction is explained by the following hypothesis: when the conduction velocity is not homogeneous along the course of the nerve because of a segmental demyelination which takes place in a spotty pattern, both longitudinally and transversely, the fastest fiber from the proximal stimulation point may not necessarily be the fastest from the distal stimulation point. If the fastest fibers from the 2 stimulating points are not the same fibers, the conduction time between these 2 points as calculated by measuring the onset of motor response may be artificially prolonged. This will give a falsely slow conduction velocity. Using this hypothesis in segmental demyelinating disease, latency alone is a better parameter of nerve conductivity than nerve conduction velocity, and serial latencies may correlate better with the clinical course than do serial conduction velocities.