Abstract
Thr correlation between 4 nerve conduction velocities (NCVs) and 9 clinical laboratory findings in 57 diabetics was studied by multivariate analysis. F-wave conduction velocity (FCV) was measured as the motor nerve conduction velocity between the spinal cord and elbow. M-wave conduction velocity was measured as that between the elbow and wrist. Proximal and distal sensory nerve conduction velocities were measured between the elbow and wrist and between the wrist and finger, respectively. The 9 clinical laboratory findings were age, sex, duration of diabetes mellitus, control of blood sugar, retinopathy, proteinuria, patellar tendon reflex, Achilles tendon reflex, and kinds of therapy. In simple correlation analysis, the 4 NCVs were found to be closely related to control, retinopathy and kinds of therapy. However, in partial correlation analysis, only FCV revealed a close and positive correlation with good control of blood sugar and a negative correlation with age. The F-wave is caused by excitation of anterior horn cells of the spinal cord by antidromic electric stimulation of motor fibers. The function of the motor fibers between the spinal cord and elbow had a positive correlation with good control of blood sugar.
Factor analysis of the 4 NCVs involved division into two factors. The first factor was closely related to age. The second was closely related to control and therapy in partial correlation analysis. FCV was estimated by multiple regression analysis from the 9 clinical laboratory findings. The multiple correlation coefficient was 0.753. The percentages of the estimated FCV values which were within a maximum error of 2 m/sec and 3 m/sec of the measured FCV were 53 and 72, respectively.
Multivariate analysis is thus useful for estimating the relationship between NCVs and control of diabetes mellitus.
