Screening for type 1 diabetes – when, who and how?

People at risk of type 1 diabetes can be identified before symptoms occur, reducing the risk of an emergency diagnosis. Initiatives around the world looking at population-based screening for the condition were discussed at the recent Advanced Technologies & Treatments for Diabetes (ATTD) conference. Dr Susan Aldridge reports.

Type 1 diabetes does not arrive ‘out of the blue’, but goes through three well-described stages that culminate in clinical diagnosis. Stage 1 is marked by the presence of two or more islet-specific antibodies, but normoglycaemia. Stage 2 is characterised by dysglycaemia and a fall in beta cell mass, but is still asymptomatic. Stage 3 is clinical type 1 diabetes. 

Unfortunately, type 1 diabetes all too often presents as an emergency which may involve diabetic ketoacidosis (DKA). “I think of this as a plane or train crash, whether or not it’s associated with DKA,” said Dr Rachel Besser, consultant paediatric diabetologist at the Oxford NIHR Biomedical Research Facility. “It causes huge psychological trauma for families because a few die from late diagnosis.” This happened to Peter Baldwin, aged 13, in 2015 after failure to diagnose type 1 diabetes when he became ill with DKA. His family has since been campaigning for the introduction of screening which, said Dr Besser, is the right approach because type 1 diabetes can then be identified at an earlier stage before someone gets symptoms. 

According to recent research, DKA rates have been increasing, even before COVID-19, and the acidic milieu it produces is bad for the developing brain and pancreas in children. It decreases IQ and is associated with deteriorating HbA1c. It seems that public awareness has been insufficient to prevent DKA. “So there are two reasons to screen – to diagnose early and prevent DKA, so that diagnosis is a ‘smooth landing’ and not a crash, and to start prevention of clinical diagnosis with teplizumab,” said Dr Besser.  

Population screening studies

Screening for type 1 diabetes started among families because first-degree relatives of someone with type 1 diabetes have a 15 times higher risk of developing the condition than the general population (5% versus 0.3%). However, this approach only captures a minority of those who are in the early stages, so there is still a need to screen more general population groups. This can be done by screening for antibodies alone or in conjunction with genetics.

For instance, the TEDDY study screened more than 400,000 people and identified more than 8,000 genetically at risk through their human leukocyte antigen (HLA) status, and followed them up for 15 years, providing a wealth of data. Non-HLA loci can also be used and combined with HLA to give a genetic risk score (GRS) and thereby identify 10% of those at risk of having multiple antibodies compared with just 5% using only HLA status. This approach can halve the cost of early-life intervention and over 1,000 babies around the world have been recruited into an oral insulin prevention trial thanks to this work. 

Combining GRS with islet antibody testing is being used in the CASCADE trial, which is also looking at coeliac disease as part of a newborn screening programme, with a target of screening 60,000 babies. “I like this approach because they’re using electronic health records to find out who is eligible for the trial and also to flag up to healthcare providers when a child is at risk,” said Dr Besser. There is also the Type 1 Diabetes National Screening Pilot in Australia, funded by the JDRF, which aims to find the best way of screening for type 1 diabetes in children. 

To this end, they are using three different approaches – in the first two, GRS and antibodies are used, the only difference being that one group has both done via a heelprick test at birth, while the other has a saliva swab between six and 12 months for the GRS, then an antibody test. The third approach involves screening just for islet antibodies on a dried blood spot. Meanwhile, JDRF has its TIDetect screening awareness and education programme, which advises on why screening is important, where to get screened and what to do with the results.  

When to screen

“In my view, type 1 diabetes screening needs to be integrated into the healthcare system,” said Dr Besser. As for what age type 1 diabetes screening should start, insulin antibody peaks first, followed by the other islet antibodies, and the peak age for developing islet antibodies is between nine months and two years. “If you did one test between the age of three and five, you’d miss the younger kids and a child under the age of three has a very high risk of DKA and a very vulnerable brain – we don’t want to miss them,” said Dr Besser, “but then you might miss the older children whose antibodies appear later.”  

This means that screening at two different ages might be helpful. So thinking about a targeted approach that can hopefully be integrated better into a healthcare system, there are lessons to be learned from the Fr1da study in Bavaria, involving more than 600 primary care paediatricians. From 2015, they have done one screen from age two to six years and then, from 2020, two screens between the ages of two and 10 years, as part of a ‘well child’ check. 

“They have paved the way in telling us so much about what we can learn from screening,” said Dr Besser. So far, more than 170,000 children have been screened and 483 confirmed with multiple islet antibodies and followed up. Rates of DKA have in the region have improved as a result and, with earlier diagnosis, presenting HbA1c is much lower. Similar work is going on with the Antibody Detection Israeli Research Program, which aims to screen the paediatric general population at two ages – nine to 18 months and two to five years. 

Meanwhile, Dr Besser is leading on a type 1 diabetes early-detection programme in Oxford, where screening is being done at the pre-school vaccination stage (aged three to five years) with antibody testing.  

Going forward

For developing screening programmes, there are guidelines, such as those developed by Wilson and Jungner, which are already widely used. And, besides clarifying ages for screening, implementation into routine labs is needed, as is validation of test performance. Also the value of a single antibody result is not yet fully established, while more data is still needed on the benefits and risks of type 1 diabetes screening.  

So what should be done for someone who tests positive for two or more antibodies? The International Society for Pediatric and Adolescent Diabetes has developed guidelines from a consensus group that can hopefully be used in a clinical practice within the next 12 months. Dr Besser concluded by sharing a case history of a three-year-old child who had started on insulin as an outpatient – no hospital admission, no DKA, so very different from the tragic case of Peter Baldwin and others. “It’s a new face for type 1 diabetes and I think it’s here to stay,” she said. “The goals are early detection before DKA sets in and recruitment into clinical trials. We need to figure out how we’re going to screen – genetics with antibodies or antibodies alone, with or without other conditions, venous, capillary or dried blood? What are we going to do about follow-up and, of course, all of this has implications for the population in your country and country-specific costs. We want a smooth landing, not a plane crash, into type 1 diabetes diagnosis.”  

To learn more, enrol on the EASD e-Learning course ‘Diagnosis of type 1 diabetes’.

Any opinions expressed in this article are the responsibility of the EASD e-Learning Programme Director, Dr Eleanor D Kennedy.