Assessment of Two Different Glucagon Assays in Healthy Individuals and Type 1 and Type 2 Diabetes Patients

Abstract

Methods for glucagon analysis suffered in the past from lack of specificity and a narrow sensitivity range, which has led to inaccurate results and to the suggestion that type 1 diabetes (T1D) and type 2 diabetes (T2D) patients have elevated fasting glucagon levels. However, the availability of more specific and more sensitive methods to detect intact glucagon has shown that actual glucagon levels are lower than previously assumed. This study aimed to characterize fasting plasma glucagon levels in healthy individuals and T1D and T2D patients with two different glucagon assays. The study included 20 healthy individuals, 20 T1D and 20 T2D patients. Blood was collected under fasting conditions. A double-antibody sandwich enzyme-linked immunosorbent assay (ELISA) and a conventional radioimmunoassay (RIA) were used. A significant difference in fasting glucagon levels between healthy individuals and T2D was observed by ELISA, but not by RIA. ELISA also yielded lower glucagon levels in healthy individuals than in T1D and T2D patients which RIA did not. RIA produced significantly (p = 0.0001) higher overall median glucagon values than ELISA in a pooled analysis. These results underline the notion that the choice of selective laboratory methods is highly relevant for mechanistic endocrine research.

Keywords: ELISA; RIA; glucagon; glucagon assay; type 1 diabetes mellitus; type 2 diabetes mellitus.

Figure 1

Glucagon levels in healthy individuals, T1D and T2D patients assessed with ELISA and RIA. * p = 0.005 (A); * p < 0.001 (B).

Figure 2

Bland–Altman plot. x-axis: The average (pmol/L) of ELISA and RIA for each plasma sample; y-axis: The mean difference (pmol/L) of ELISA and RIA. The solid line indicates the mean difference (bias) between the methods. The dashed line and the upper dotted line indicate the limits of agreement between the two methods (95% limits of agreement).