Reproducibility and variability in the action of injected insulin

Abstract

Insulin-treated patients are generally taught to adapt their doses of insulin according to the glycemic level obtained during self-tests. They usually adhere to medical recommendations, but are often confused by the results, which may not correspond to expectations. Patients have to contend with variability and a certain degree of unpredictability in the results. Our knowledge of the factors involved in this variability is often imprecise. We review here the factors depending on the preparation of insulin itself, not only with regard to its crystallization but also the speed at which the hexamers dissociate into dimers. The development of fast and slow-acting analogues is discussed along with their value in improving glycemic predictability. In addition to these factors, we mention those stemming from the injection technique itself, which are directly related to the instructions given to the patients. For crystallized insulin preparations, shaking the bottle is an important element that the development of slow-acting analogues should eliminate, but the time lapse before withdrawing the needle, the anatomical site of the insulin injection, and the depth of the injection are also factors for variability. Greater predictability in the action of insulin will be obtained from a combination of progress in manufacturing procedures and better patient education.