Abstract
It is clinically important to diagnose diabetic neuropathy at the early stage. In this study, the toneentropy analysis of electrocardiogram was applied to diabetic patients to evaluate its usability in the screening of diabetic neuropathy. Consecutive 102 diabetic patients were invited to the study. Electrocardiogram was obtained and analyzed for the tone and the entropy using an original software developed previously. Nerve conduction velocity (NCV) was examined on the median, the posterior tibial and the sural nerves. Patients were divided into quartile subpopulations according to the NCVs in the analysis. Both the tone and the entropy significantly correlated with NCVs, while coefficient of variation in R-R intervals did not show a significant correlation. The correlation was most significant between the entropy and the NCV on the sural nerve. When a multivariate analysis (ordinary regression) was applied to examine independent effects of the factors influencing the NCV on the sural nerve, the entropy was the most potent independent factor (β = 1.14 ± 0.32, P = 0.0004) along with sex (β = 0.43 ± 0.19, P = 0.02) and BMI (β = 0.11 ± 0.05, P = 0.04). The tone-entropy analysis on electrocardiogram may be a promising non-invasive screening method for diabetic neuropathy.