In two studies of patients with diabetes who did not have neurologic symptoms, nerve conduction velocity was increased either by an improvement in glucose control or by the administration of the aldose reductase inhibitor sorbinil. In a 1981 study by Graf et al, glycemic control and motor and sensory nerve conduction velocities were evaluated in 18 patients with non-insulin-dependent diabetes before and after one, three, six, and 12 months of antihyperglycemic therapy. There was an improvement in motor nerve conduction velocity (median motor, p less than 0.01; peroneal motor, p less than 0.05; and tibial motor, p less than 0.05), which was associated with the improvement in fasting plasma glucose levels after three months for some motor nerves (median motor: r = -0.62, p less than 0.01; peroneal motor: r = -0.50, p less than 0.05). A direct linear relationship between the change in fasting glucose and glycosylated hemoglobin levels and the change in median motor nerve conduction velocity after 12 months of antihyperglycemic therapy was also found. Thus, there was a tendency for those patients who had the greatest improvement in glycemic control to have the greatest improvement in motor nerve conduction velocity. The findings in the first study are consistent with the hypothesis that hyperglycemia contributes to slowed nerve conduction velocity. In a 1983 randomized, double-blind, crossover study by Judzewitsch et al, motor and sensory nerve conduction velocities were evaluated in 39 patients with insulin-dependent or non-insulin-dependent diabetes in whom glycemic control was stable. During nine weeks of treatment with 250 mg per day of sorbinil, nerve conduction velocity was faster in the three tested nerves when compared with the velocities during the placebo period (peroneal motor nerve conduction velocity: +0.70 +/- 0.24 m per second, mean +/- SEM, p less than 0.008; median motor nerve conduction velocity: +0.66 +/- 0.27 m per second, p less than 0.005; median sensory nerve conduction velocity: +1.16 +/- 0.50 m per second, p less than 0.035). Although the effect of an improvement in glycemic control and administration of sorbinil in increasing nerve conduction velocity in two groups of neurologically asymptomatic patients with diabetes was small, the findings are consistent with the hypothesis that polyol pathway activity contributes to slowed large-fiber nerve conduction velocity in these patients.