Precipitation of insulin by alcohol and some aspects of the "back-titration" method of insulin immunoassay were investigated. In the absence of insulin antibodies, alcohol has a definite nonspecific precipitant action on labeled insulin, an action which is most marked when alcohol is added in high concentration or when the medium contains much protein. This nonspecific precipitation of radio-iodinated insulin is not affected by the presence of high concentrations of unlabeled insulin; this is

more »

... t explained but has no effect upon the assay system. The "back-titration" method of insulin assay is more sensitive and reproducible than the "competitive" technic when the latter is carried out under similar brief conditions of incubation. The "back-titration" method is shown to be specific for insulin, proinsulin being the only other hormonal substance tested and found to interfere with the assay system. The assay is not affected by small changes in pH about neutrality, can be carried out with pooled anti-porcine or anti-bovine insulin serum, and is not species-specific. Results obtained with human plasma are comparable with those found by the method of Morgan and Lazarow. Expected amounts of insulin are recovered over a wide range of plasma dilutions. For routine purposes, illustrated in a series of glucose tolerance tests, the "backtitration" method has proved practical, rapid, and reproducible. DIABETES 20: 33-45, January, 1971. Methods currently used for immunoassay of insulin differ from one another in one major respect. Separation of antibody-bound from free labeled insulin in final reaction mixtures is accomplished by a variety of technics including chromato-electrophoresis, 3 precipitation of antibody-bound hormone with sodium sulfite 7 or a second antibody,