A screening test looks for the immune attack on the islets before the person has any sign of it. The attack leaves a trace in the blood as autoantibodies, and that trace can appear while glucose is still normal, as the natural-history note showed. So a test for autoantibodies looks at the process itself rather than at glucose. Finding the trace is part of the point. A positive result is worth having if it changes what happens next, and that question is what the screen hangs on.
Can the attack be seen before any symptom?
Islet autoimmunity is the immune attack on the insulin-producing cells of the pancreatic islets. It can be detected by autoantibodies before clinical type 1 diabetes (T1D) appears. This is what makes screening possible. The moment an autoantibody first becomes detectable in the blood is called seroconversion.
Which trace does the test look for?
That trace is a set of antibodies, and the screen looks for five of them. Five known autoantibodies can be screened for in a population to identify people at risk of progressing to T1D:
- islet cell antibodies (ICA)
- autoantibody to insulin (IAA)
- antibodies to tyrosine phosphatase (IA-2)
- antibodies to glutamic acid decarboxylase (GAD)
- antibodies to a zinc transporter (ZnT8)
Does a positive result mean the disease will come?
If the antibodies are there, the next question is what they predict. Nearly all people who develop seroconversion for islet autoimmunity eventually progress to clinical T1D. A person with two or more autoantibodies and normal or impaired glucose tolerance is in stage 1 or stage 2 disease, and the natural-history note sets out how those stages are defined.
Can the change be managed without a crisis?
Knowing the stage is useful if the change can be treated before a crisis. The crisis here is ketoacidosis, which usually follows when insulin deficiency is accompanied by a rise in counter-regulatory hormones. A person at stage 1 or stage 2 still has beta cells and is starting to drift towards that deficiency. So the person can be monitored, and the transition to clinical disease can be recognised and treated without a crisis. The value of screening lies in what the result allows.
Is there a treatment that changes the course?
If a treatment can act at stage 2, the next question is who can be found at that stage. At stage 2 there is now also a treatment that changes the course of the disease: teplizumab. It was approved in 2022 as the first disease-modifying therapy for T1D.
Its effect applies in people with stage 2 disease. It delays the onset of stage 3 disease by an average of about two years. It does not prevent clinical diabetes.
Why not screen everyone, or only relatives?
Treatment and monitoring help if the right people are found, so the question becomes who to look at. Screening is still under trial and is not applied at the level of the whole population. More than 85% of T1D cases have no family history of the disease, as the pathogenesis note showed. Screening only relatives would therefore miss most cases of T1D.
How do the trials choose whom to test?
The trials start with genes and then test the blood at set ages. They begin with genetic testing at birth for the human leukocyte antigen (HLA) genes, the immune-system genes that carry the main inherited risk.
After the genetic test, participants are tested for autoantibodies twice, at ages 2 and 4 years. The trials are still investigating the best cost-effective way to screen.
What happens outside the trials?
Outside these trials, type 1 diabetes is generally recognised only when symptoms appear. That fits what the natural-history note showed: glucose stays normal while enough beta cells remain to keep up, and rises as the reserve shrinks. So what do those symptoms look like, and why can they arrive as an emergency?