Diagnostic sensitivity of predicted F-wave latency by age, height, and MCV.

Abstract

The aim was to investigate whether minimum F-wave latency (Fmin) predicted by a multiple regression equation utilizing age, height, and motor nerve conduction velocity (MCV) (3-factor method) was clinically useful to detect Fmin abnormality in diabetic polyneuropathy (DP) and lumbosacral monoradiculopathy (RAD). Thirty-one DP patients (ulnar, tibial, and peroneal nerves) and 66 RAD patients (peroneal nerve) were investigated. The specificity and sensitivity of the 3-factor method were compared to those obtained by a univariate (height) regression method (1-factor method) and another multivariate (height and age) regression method (2-factor method). In general satisfactory specificity was obtained with all three methods. In DP patients the tibial 3-factor method showed statistically lower sensitivity than the other two methods. There were no significant differences in diagnostic sensitivity among the three methods in two other nerves of the DP patients and the peroneal nerve of the RAD patients. The 3-factor method was useful to estimate Fmin in normal subjects, but it was of little value in increasing the diagnostic sensitivity in DP and RAD patients. The 2-factor method was more appropriate for detecting DP.