Because of its strong efficacy, long-term durability, and predictability when titrated with algorithms employed in clinical trials, basal insulin remains a mainstay of treatment for type 2 diabetic patients who are not able to attain or maintain glycemic control with oral agents alone. Because diabetes is progressive disorder, even patients who are initially well controlled on basal insulin will likely require “fine tuning” of their diabetes regimen, usually with agents that preferentially affect postprandial glucose levels.

GLP-1 agonists (e.g., albiglutide, dulaglutide, exenatide, liraglutide) can impact postprandial glucose levels by means of blunting glucagon release, as well as modulating gastric emptying. In patients who have achieved fasting glucose control on basal insulin — when postprandial glucose is the target — would prandial insulin or GLP-1 agonist better serve their needs?

Digenio et al report on a “real world” data analysis that looked retrospectively at a large population of General Electric Centricity electronic health records (n = 33,848) from persons who had either prandial insulin or a GLP-1 agonist added to basal insulin when glucose control was not adequate.

At follow-up 6 months-1 year later, GLP-1 agonists and prandial insulin provided similar A1c reductions, but there were substantial differences in weight (gain with rapid acting insulin vs. loss with GLP-1 agonist) and hypoglycemia (more frequent with insulin).

Current ADA/EASD guidelines include weight gain and risk of hypoglycemia as critical considerations for advancement of therapy. These data support the concept that in “real world settings,” GLP-1 agonists perform essentially as well as rapid acting insulin, with less hypoglycemia and weight loss instead of gain.