Morphometric analysis of horizontal canal nerves of chronically diabetic rats.

Horizontal canal nerves of 3-, 6-, 9-, and 12-month-diabetic rats were compared with those of age-matched controls. The myelin sheaths of the horizontal canal nerves in diabetic rats were thinner than those of age-matched controls (mean +/- SD 0.63 +/- 0.04 micron (n = 16) vs. 0.71 +/- 0.05 micron (n = 9); p < 0.0001, one-tailed t test). Regression analysis revealed that myelin sheath thickness did not correlate with severity of diabetes, but myelin thinning did occur as a function of the duration of diabetes (p < 0.05, regression ANOVA). The progression of myelin thinning over time is consistent with the possibility of an accelerated decline in vestibular function with age in diabetic patients. That myelin thinning did not correlate with the severity of diabetes suggests that this thinning is not directly related to the aging effects attributed to nonenzymatic glycosylation of myelin proteins. Multivariate analysis revealed a significant difference between diabetic and control groups when fiber diameter and intrasheath diameter were considered together (p < 0.008, canonical discriminant-function analysis). Diabetic and control groups did not differ significantly in total nerve fiber counts. In the diabetic group, however, nerve fiber counts were higher in animals with higher blood glucose levels (p < 0.02, linear-regression ANOVA; r2 = 0.49). The finding of higher nerve fiber counts in more severely diabetic rats is consistent with an earlier transmission electron microscopic finding of false myelinated nerve fiber profiles in micrographs from more severely diabetic rats. These false profiles are believed to represent phagocytosis-like Schwann cell reactions against their own myelin, triggered by excess myelin glycosylation.