Salivary insulin in normal and type I diabetic subjects.

We studied the relationship of salivary insulin to serum insulin concentrations in normal subjects and type I (insulin-dependent) diabetic patients to test the hypothesis that salivary insulin might be a simple measure of insulinemia in diabetes. In 8 nondiabetic subjects, salivary insulin levels increased after an oral glucose load but with a delay in peak concentrations of approximately 45 min in comparison with serum insulin levels. There was a significant correlation (r = .810, P less than .01) between mean serum insulin and the salivary insulin 30 min later. In 12 type I diabetic patients, day profiles of saliva and serum insulin were obtained during usual insulin treatment, diet, and physical activity. In serum, the mean (+/- SE) percentage of bound insulin was 58.8 +/- 5.2%, and in saliva it was 45 +/- 3.5%. The mean ratio of salivary to serum free insulin throughout the day was 1:1.6. Although there was a significant correlation (r = .913, P less than .001) between mean serum free insulin for all patients and the corresponding mean free salivary insulin, several individual profiles showed marked discrepancies between the timing and magnitude of insulin changes in the two compartments. We would not, therefore, recommend salivary insulin concentrations as a reliable index of insulinemia in individuals with type I diabetes.