GLP-1s and Type 1 Diabetes

I am writing this because people with type 1 diabetes ask and get either a flat no or an enthusiastic yes, and neither is the honest answer.

No GLP-1 receptor agonist is approved for type 1 diabetes. Use in this setting is off-label, professional guidance does not recommend it routinely, and it belongs with an endocrinology team rather than with a weight-management service or a telehealth form.

That said, the reasons people ask are legitimate.

Why the question arises

Type 1 diabetes and obesity increasingly coexist, with excess weight now affecting a large share of adults with type 1 — sometimes called double diabetes, meaning autoimmune insulin deficiency plus acquired insulin resistance.

The loop is difficult. Insulin is a storage hormone, so intensive therapy promotes weight gain. Weight gain raises insulin requirements. Higher doses promote further gain. And the obvious response — eating less — carries a risk in someone whose insulin is injected rather than regulated, because a reduced intake against a fixed dose is a hypoglycemia problem.

Adults with type 1 also develop the same cardiovascular, hepatic and joint consequences of excess weight as everybody else. They are not exempt from the reasons these drugs exist.

What the trials showed

Liraglutide added to insulin in type 1 diabetes was studied properly, in the ADJUNCT programme and in smaller trials such as Lira-1. The pattern across them:

• Modest HbA1c reductions, in trials running 26–52 weeks
• Modest weight reduction
• Reduced insulin dose requirements
• More hypoglycemia
• More ketosis-related events

That last pair is why the balance did not favour approval. The benefits were real and small; the risks were real and concerning.

Semaglutide and tirzepatide have not been through equivalent programmes in type 1, so their use here rests on smaller studies, observational reports and extrapolation — which should be read with the caution set out in how to read a GLP-1 study.

The risk that actually matters

Not hypoglycemia, though that matters. Diabetic ketoacidosis, and specifically the version that does not announce itself.

Euglycemic DKA is ketoacidosis occurring with a glucose that reads normal or only mildly elevated. It happens when insulin is insufficient relative to need while carbohydrate intake is low: the body burns fat, ketones accumulate, acidosis develops — and the glucose reading never rises enough to alarm anyone.

A drug whose entire purpose is to substantially reduce food intake creates exactly those conditions. Add an intercurrent illness, or an SGLT2 inhibitor, and the risk compounds further — the SGLT2 version of this problem is described in sick day rules.

The safeguard is ketone testing, not glucose testing. If you have type 1 diabetes and are taking a GLP-1, you should have ketone strips or a meter, and you should test whenever ketones may be rising — within 24 hours of any of these: when you are unwell, when you are vomiting, when you have eaten very little for a day, and whenever you feel wrong in a way a glucose reading does not explain.

Symptoms to act on: nausea and vomiting, abdominal pain, deep or rapid breathing, a fruity smell on the breath, drowsiness — with a glucose that looks acceptable.

The insulin rule

One sentence, and it is the most important in this article.

Never reduce or omit basal insulin because your appetite has gone.

Mealtime insulin is matched to the carbohydrate you actually eat, and reducing it when you eat less is expected and correct. Basal insulin is different — it covers your background metabolic requirement whether or not you eat anything at all. Cutting it because you are not hungry is the classic route into ketoacidosis, and it is an easy mistake to make on a drug that makes food feel unnecessary.

If you are unwell and not eating, you still need basal insulin. That is the same rule taught in every type 1 sick-day protocol, and it applies with more force here.

Basal changes come from your diabetes team. So do the initial reductions when starting, which are usually needed and should be planned rather than discovered — the general framework for that is in hypoglycemia on a GLP-1.

What supervision should look like

If a specialist team supports a trial of this off-label, I would want:

• Endocrinology-led, with continuous glucose monitoring in place and ketones checked within 24 hours of any illness
• A written insulin adjustment plan for starting and for each escalation
• Ketone testing supplies and explicit instructions on when to test
• A sick-day plan that names ketones, not just glucose
• Very slow titration — 8 weeks between steps rather than the standard 4 weeks
• A defined review point at 6 months, at which it is stopped if it is not clearly helping
• Attention to protein and resistance training, because lean mass loss on top of type 1 is not a good trade — muscle is the whole game

And what should stop it

Any episode of ketoacidosis. Recurrent unexplained ketosis. Severe or unaware hypoglycemia. Or an inability to maintain adequate intake, which in this group is not merely a nutritional problem but a metabolic one.

I would rather someone with type 1 diabetes read this and decide to have the conversation properly than obtain a prescription from a service that never asked which type of diabetes they have — a failure mode covered in how to choose a GLP-1 prescriber.