Objective
To determinate the reference values of residual latencies of motor nerves and to evaluate the early diagnostic value of residual latency.

Method
The subjects were 129 diabetes mellitus patients and 60 controls with no known neurological disorders. The patients were divided into two groups based on the conventional nerve conduction study: Group 1, 75 patients without neuropathy; Group 2, 54 patients with neuropathy. The group 2 patients were subdivided into 4 sub- groups on the basis of conduction velocity and residual latency of the median nerve. Residual latencies were measured in all subjects and glycosylated hemoglobin percentages (HbA1c) were measured in the diabetes patients. In group 2, each nerve conduction parameter was correlated with the duration of diabetes and HbA1c. The duration of diabetes, HbA1c, and amplitude of median nerve response were compared between the subgroups of group 2 patients.

Results
Motor residual latencies obtained from the controls were 1.93⁑0.28 msec, 1.53⁑0.24 msec, 2.46⁑0.43 msec, 2.21⁑0.53 msec in median, ulnar, deep peroneal and posterior tibial nerves, respectively. In group 1, motor residual latencies of median & deep peroneal nerves were significantly delayed compared with those of the controls. In group 2, motor residual latencies of median, ulnar, deep peroneal and posterior tibial nerves were significantly delayed more than those of the controls and group 1.

In group 2, increased HbA1c correlated to the decreased conduction velocities of median, deep peroneal, posterior tibial nerves but not to the residual latencies.

In the subgroup of group 2 (2-D), the nerve involved more distally showing lower compound muscle action potential and higher HbA1c.

Conclusion
Residual latency measurement can be a useful diagnostic method for the early detection of diabetic neuropathy.