Salivary immunoreactive insulin: a new entry in clinical chemistry?

The distribution of salivary immunoreactive insulin (S-IRI) and its relation to plasma insulin (P-IRI) and other clinical and metabolic variables were investigated in 93 nondiabetic subjects (60 males, 33 females, ages 11-70 y). S-IRI was measured by RIA, with monoiodinated insulin as labeled antigen and with standards consisting of serial dilutions in saliva of known amounts of native insulin. The sensitivity of the method was 12.1 pmol/L. The intra- and between-assay CVs were near 10%, and analytical recovery exceeded 93% for various insulin concentrations added to saliva. S-IRI, measured after overnight fast, showed a nonguassian distribution in males (skewness 1.18, kurtosis 1.41) and females (skewness 1.71, kurtosis 1.18); mean and median values were higher in males (14.3 and 12.1 nmol/min) than in females (12.1 and 7.15 nmol/min), but not significantly so. S-IRI was significantly correlated with P-IRI both in males (r = 0.75, P less than 0.0001) and females (r = 0.72, P less than 0.0001). These results, particularly the correlation between S-IRI and P-IRI, indicate the possibility of using S-IRI measurement in clinical practice.