There is no sufficient statistical bases on the diabetic neuropathy, because there has been poor definition and confusion of diagnostic criteria for diabetic neuropathy. We studied sensory nerve conduction(SNC) and scalp-recorded tibial somatosensory evoked potential(SEP) on the lower extremity of diabetic neuropathy.

One hundred and five patients, diagnosed as diabetes mellitus at internal medicine department of SNUH, were examined and results are followings.

1) Of the 105 DM patients, 83(78.1%) patients were diagnosed as diabetic neuropathy; Abnormal SNC was seen in 73(69.5%) patients and abnormal SEP was seen in 69(65.7%) patients.

2) Of the 71 patients with symptom, 95.8%(68 patients) were diagnosed as diabetic neuropathy and only 4.2%(3 patients) were normal by both methods.

3) Of the 34 patients without symptom, 21(61.8%) patients were normal by both methods and 13(38.2%) patients were diagnosed as subclinical neuropathy.

4) In above three groups, difference between SNC and scalp-recorded SEP was statistically significant by chi-square and Fisher test.

5) Of 73 patients showing abnormal SNC, 31 patients had no response and of 69 patients showing abnormal cortical SEP, only 13 patients had no response.

6) Snc and SEP showed tendency of prolonged latency, and decreased amplitude as the duration of DM and symptom of DM neuropathy increased. But the coefficients of correlation were low indicating there is no definite linear correlation.