Conduction slowing in painful versus painless diabetic neuropathy

Motor conduction slowing in diabetic distal symmetrical polyneuropathy (DSP) generally exceeds that in distal axonal polyneuropathy. Additional mechanisms secondary to axonal injury may contribute to motor conduction slowing; however, its clinical and pathophysiological significance has rarely been discussed. The purpose of this study was to evaluate the clinical and pathophysiological significance of conduction slowing in DSP. We analyzed motor conduction in 50 patients with symptomatic painful DSP and 25 patients with asymptomatic painless DSP. Motor conduction data from 40 patients with amyotrophic lateral sclerosis (ALS) were used as controls for pure axonal conduction slowing. Compound muscle action potential amplitude (CMAP), distal motor latency (DL), and conduction velocity (CV) values were converted to a percentage of the upper or lower limits of normal and then presented in square root transformation (SQRT) form. Plots of SQRT-DL and SQRT-CV against SQRT-CMAP were analyzed. Regression analysis showed that DL and CV are amplitude-dependent in both painless DSP and ALS. Changes of DL and CV in painful DSP were not amplitude-dependent except for DL in the lower extremities. Our data support the hypothesis that the mechanism of slowing is similar in both painless DSP and ALS, and that it results from the loss of large, fast-conducting fibres. Lack of amplitude-dependency in conduction slowing in painful DSP may reflect combined axonal and demyelinating changes.