Glucose Concentrations of Less Than 3.0 mmol/L (54 mg/dL) Should Be Reported in Clinical Trials: A Joint Position Statement of the American Diabetes Association and the European Association for the Study of Diabetes

The International Hypoglycaemia Study Group recommends that the frequency of detection of a glucose concentration,3.0 mmol/L (,54 mg/dL), which it considers to be clinically significant biochemical hypoglycemia, be included in reports of clinical trials of glucose-lowering drugs evaluated for the treatment of diabetes mellitus. The glycemic thresholds for symptoms of hypoglycemia and for glucose counterregulatory (including sympathoadrenal) responses to hypoglycemia, as plasma glucose concentrations fall, are not fixed in patients with insulin-, sulfonylurea-, or meglitinide (glinide)-treated diabetes. They are at higher glucose concentrations in those with poor glycemic control and at lower glucose concentrations in those with tight glycemic control (1–5). The shifts in glycemic threshold to lower glucose concentrations are largely the result of more frequent episodes of iatrogenic hypoglycemia during intensive glycemic therapy. Glycemic thresholds for responses to hypoglycemia vary, not only among individuals with diabetes but also in the same individual with diabetes as a function of their HbA1c levels and hypoglycemic experience; it is therefore not appropriate to cite a specific glucose concentration that defines hypoglycemia in diabetes. As a consequence, the American Diabetes Association has defined hypoglycemia in diabetes nonnumerically as “all episodes of an abnormally low plasma glucose concentration that expose the individual to potential harm” (6,7). Nonetheless, the International Hypoglycaemia Study Group believes that it is important to identify and record a level of hypoglycemia that needs to be avoided because of its immediate and long-term danger to the individual. A single glucose level should be agreed to that has serious clinical and health-economic consequences. This would enable the diabetes and regulatory communities to compare the effectiveness of interventions in reducing hypoglycemia, be they pharmacological, technological, or educational. It would also permit the use of meta-analysis as a statistical tool to increase power when comparing interventions. In its discussion, the International Hypoglycaemia Study Group considered glucose concentration levels of ,3.0 mmol/L (,54 mg/dL) and ,2.8 mmol/L (,50 mg/dL) detected by self-monitoring of plasma glucose, continuous glucose monitoring (for at least 20 minutes), or a laboratory measurement of plasma glucose. Both of these levels are distinctly low glucose concentrations that do not occur under physiological conditions in nondiabetic individuals (8). Thus, they are unequivocally hypoglycemic values. They approximate the upper and lower limits, respectively, of the nondiabetic glycemic threshold for symptoms of insulin-induced hypoglycemia (8–10). The generic nondiabetic glycemic threshold for impairment of cognitive Corresponding author: Simon R. Heller, s.heller@ sheffield.ac.uk.