For people with type 2 diabetes who already use long-acting insulin, adding a popular GLP-1 drug may not make it any easier to stop insulin.
A large study of U.S. veterans found that insulin discontinuation was about as common with GLP-1 medicines as with two other major types of diabetes drugs.
GLP-1 receptor agonists have attracted enormous attention in recent years. Medicines in this family can lower blood sugar, help many people lose weight and, for some patients, reduce the risk of serious heart and kidney problems.
Because these drugs can improve blood sugar control and reduce the amount of insulin some people need, doctors and patients may wonder whether starting one could eventually allow insulin treatment to be stopped. The new research suggests that this does happen for some patients, but not more often than after starting other glucose-lowering medicines.
The study was led by Kasia J. Lipska, MD, of Yale School of Medicine and included veterans with type 2 diabetes who were receiving basal insulin. Basal insulin is a long-acting form of insulin designed to provide a steady background level throughout the day and night.
Type 2 diabetes develops when the body cannot keep blood sugar within a healthy range. Many people initially manage the condition with lifestyle changes and non-insulin medicines, but some eventually need insulin because their bodies cannot make enough of the hormone to control blood sugar.
The researchers compared three groups of commonly used diabetes medicines. These were GLP-1 receptor agonists, SGLT2 inhibitors and DPP-4 inhibitors.
These drug families lower blood sugar in different ways. GLP-1 drugs help the body release insulin when needed, slow digestion and reduce appetite, while SGLT2 inhibitors cause the kidneys to remove more glucose through urine.
DPP-4 inhibitors work by helping natural hormones involved in blood sugar control remain active for longer. Unlike GLP-1 drugs, they generally have a smaller effect on body weight.
The researchers examined U.S. veterans who started one of these medicines between 2020 and 2022 while already using basal insulin. They created 8,869 matched sets of patients so that people starting the different treatments could be compared as fairly as possible.
The study used a method called a target trial emulation. Instead of randomly assigning people to medicines in a new clinical trial, researchers used existing health records and statistical methods to imitate some of the features of a randomized trial.
Over three years, 16.7% of patients who started a GLP-1 drug stopped insulin. The figure was 17.9% among those starting an SGLT2 inhibitor and 17.1% among people starting a DPP-4 inhibitor.
These small differences were not strong enough to show that GLP-1 drugs provided a meaningful advantage. Additional analyses that considered whether patients continued following their original treatment strategy produced broadly similar results.
The researchers also examined different groups of patients to see whether GLP-1 drugs worked better for certain people when the goal was stopping insulin. They found no subgroup in which GLP-1 treatment clearly increased the likelihood of insulin discontinuation compared with the other medicines.
That does not mean GLP-1 drugs are ineffective or unimportant. Their value goes well beyond whether a patient can stop insulin, and certain GLP-1 medicines have demonstrated benefits involving weight, cardiovascular health and kidney outcomes.
It also does not mean that no individual patient can reduce or stop insulin after beginning a GLP-1 drug. Diabetes treatment is highly individual, and changes in weight, diet, physical activity, remaining insulin production and other medicines can all affect how much insulin a person needs.
The study instead answers a narrower question. Among veterans already receiving basal insulin, simply choosing a GLP-1 drug rather than an SGLT2 or DPP-4 medicine did not make insulin discontinuation more likely over the following three years.
The research was published online July 14 in the Annals of Internal Medicine. It provides useful real-world evidence for doctors and patients deciding what they should expect when another medicine is added to basal insulin treatment.
There are important limits to the findings. The study relied on observational health data rather than directly assigning treatments at random, so even careful matching cannot remove every possible difference between the groups.
The veteran population may also differ from the broader population with type 2 diabetes in age, sex, health conditions and other characteristics. This means the exact results may not apply equally to every patient.
Overall, the study challenges the idea that starting a GLP-1 medicine will commonly provide a special route away from insulin. These drugs remain powerful treatment options, but their benefits should be judged across blood sugar control, weight, heart and kidney health rather than by insulin discontinuation alone.
Patients should never stop or sharply reduce insulin without medical guidance. If insulin requirements change after a new diabetes medicine is started, doses should be adjusted with a healthcare professional to avoid dangerously high or low blood sugar.
Source: Yale School of Medicine.
