Comparison of three different methods for the quantification of equine insulin

T. Warnken, K. Huber, K. Feige

Exact analysis of equine insulin in blood samples is the key element for assessing insulin resistance or insulin dysregulation in horses. However, previous studies indicated marked differences in insulin concentrations obtained from sample analyses with different immunoassays. Most assays used in veterinary medicine are originally designed for use in human diagnostics and are based on antibodies directed against human insulin, although amino acid sequences between equine and human insulin differ. Species-specific assays are being used more frequently and seem to provide advantages compared to human-specific assays. The aim of this study was to compare three immunoassays, one porcine-specific insulin enzyme-linked immunosorbent assay (ELISA), advertised to be specific for equine insulin, one porcine-specific insulin radioimmunoassay (RIA) and one human-specific insulin chemiluminescence immunoassay (CLIA), all three widely used in veterinary laboratories for the analysis of equine insulin. Furthermore, we tested their clinical applicability in assessing insulin resistance and dysregulation by analysis of basal blood and blood samples obtained during a dynamic diagnostic stimulation test (OGT) with elevated insulin concentrations. Insulin values obtained from the ELISA, RIA and CLIA, investigated for analyses of basal blood samples differed significantly between all three assays. Analyses of samples obtained during dynamic diagnostic stimulation testing with consecutively higher insulin concentrations revealed significantly (p < 0.001) lower insulin concentrations supplied by the CLIA compared to the ELISA. However, values measured by ELISA were intermediate and not different to those measured by RIA. Calculated recovery upon dilution, as a marker for assay accuracy in diluted samples, was 98 ± 4 % for ELISA, 160 ± 41 % for RIA and 101 ± 11 % for CLIA. Our results indicate that insulin concentrations of one sample measured by different methods vary greatly and should be interpreted carefully. Consideration of the immunoassay method and reliable assay-specific reference ranges are of particular importance especially in clinical cases where small changes in insulin levels can cause false classification in terms of insulin sensitivity of horses and ponies.