Assay-dependent variability of serum insulin levels during oral glucose tolerance test: influence on reference intervals for insulin and on cut-off values for insulin sensitivity indices.

BACKGROUND: The oral glucose tolerance test (oGTT) is the most common method to estimate indices of insulin sensitivity in clinical as well as in epidemiological studies. The aim of this study was to investigate the variability of insulin levels for reference intervals and of insulin sensitivity indices during oGTT by three different insulin immunoassays.
METHODS: Serum insulin levels were determined during oGTT of randomly selected adult subjects (total n=101, with n=68 with normal glucose tolerance) by the LIAISON (LIA), the Elecsys (EL) and the AutoDELFIA (AD) method.
RESULTS: Sensitivity and precision of all three insulin assays were comparable. Insulin levels demonstrated a close (p<0.001) interassay correlation (LIA vs. EL: r=0.987, LIA vs. AD: r=0.966, EL vs. AD: r=0.965). Insulin levels and insulin sensitivity indices (n=68) of the AD method were significantly different compared to the LIA or EL assay (p<0.001). Basal insulin levels demonstrated an assay-dependent significant body mass index (BMI) dependency (LIA: r=0.24, p=0.05; AD: r=0.44, p<0.001; EL: r=0.36, p<0.01) in the regression analysis. At 120 min post glucose, gender was the relevant influencing factor on insulin levels of the LIA (r=0.40, p<0.001), AD (r=0.37, p<0.01) and EL (r=0.40 p<0.001) method.
CONCLUSIONS: We conclude that assay-dependent reference intervals should be used for the clinical interpretation of insulin levels from oGTT and of insulin sensitivity indices. Moreover, BMI and gender have to be considered as influencing factors.