BACKGROUND
Neuropathy is the most common complication of diabetes mellitus but the prevalence is variable because there are many neurologic tests and criteria for diagnosis of neuropathy. Subjective symptoms of patients alone are not accurate for diagnosis of neuropathy, so objective sensory tests must be added. There have been few studies about correlation between each neurologic tests; furthermore there have been no studies on comparisons between each neurologic test in Korean diabetic patients. METHODS: From September 1997 to June 1998, 142 type 2 diabetic patients visited Kyungpook National University Hospital and included in this study. Every patient had nerve conduction study and vibration threshold test, pressure threshold test of 1 point and 2 point touch stimuli by pressure specified sensory device. Some of these patients had 10g monofilament test. From receiver operating characteristics curve, sensitivities of each test were calculated and compared with each test. RESULTS: 1. From receiver operating characteristics curve, when specificities of each test reached 90%, sensitivity of vibration perception threshold test of left great toe was 69.2 %, one point discrimination touch threshold test of left great toe was 50 %, and two point test was 31.6 %. Sensitivity of 10 g monofilament test was 57% and specificity was 85%. 2. Comparisons between each component of nerve conduction study and other neurologic test showed conduction velocity and amplitude were highly correlated with vibration perception threshold test, but F wave latency was highly correlated with one point discrimination threshold test. CONCLUSION: We though vibration perception threshold, lOg monofilament and quantitative sensory tests is adequate, simple and convenient tests in order to diagnose peripheral polyneuropathy. Of all these tests, vibration perception threshold was most useful as well as a sensitive test for screening and diagnosis of peripheral polyneuropathy in nan-insulin dependent diabetic patients.