Diabetes Screening for Patients on Long-Term Steroid Therapy

Glucocorticoids are widely prescribed in the treatment of many inflammatory, autoimmune or allergic diseases, but also in transplant patients or patients with malignant hematological diseases (leukaemia, lymphoma, myeloma, etc.). They have many side effects, including effects on carbohydrate metabolism. Corticosteroids are diabetogenic because they disrupt pancreatic insulin secretion and decrease the insulin sensitivity of target tissues (liver, muscle, adipose tissue). The use of glucocorticoids may lead to the development of corticosteroid-induced diabetes in non-diabetic patients. Corticosteroid-induced diabetes is frequently found in clinical practice. Incidence figures for corticosteroid-induced diabetes vary widely. Case-control studies show odds ratios of corticosteroid-induced diabetes onset ranging from 1.36 and 2.23. In two of these studies, the diagnostic criterion was the prescription of a hypoglycemic treatment. Smaller studies, retrospective or prospective, show incidence rates of corticosteroid-induced diabetes ranging from 8.8 to 52%. This difference in incidence rates is explained by the different glucocorticoid doses, glucocorticoid durations, conditions requiring glucocorticoid therapy, and diagnostic criteria for corticosteroid-induced diabetes that vary from study to study. In kidney transplantation, diabetes concerns 16% of patients before the age of 40, and up to 52% after the age of 50. The diagnostic criteria used in the literature are often fasting and post prandial glycemia. The differences in incidence rates found with these two diagnostic methods can be explained by the pharmacokinetics of glucocorticoids. The main glucocorticoids prescribed have a duration of action of 12 to 16 hours. As they are generally prescribed in the morning as a single daily dose, fasting morning blood glucose levels are often normal while postprandial blood glucose levels are increased. This observation has been presented in 2 studies. However, the 2 diagnostic criteria were not statistically compared in these studies. While according to the American Diabetes Association, the 3 criteria used to diagnose diabetes are fasting glycemia, glycemia 2 hours after an oral glucose load of 75 g and glycated hemoglobin, these last 2 diagnostic criteria are, to the investigator's knowledge, not used in the scientific literature to screen for corticosteroid-induced diabetes. In current practice, the diagnostic criteria used are disparate. To date, there are no studies that have determined which diagnostic criterion is the most effective in screening for corticosteroid-induced diabetes. The current recommendations recommend early diagnosis and treatment in order to achieve satisfactory glycemic control as quickly as possible. These recommendations have been shown to be effective in reducing the risk of degenerative complications of diabetes, but it is important to screen for corticosteroid-induced diabetes with the most relevant diagnostic criterion in order to optimise its management.