Misdiagnosis of Adult-Onset Type 1 Diabetes Is Common

Miriam E. Tucker

October 18, 2018

BERLIN — Onset of type 1 diabetes after age 30 years is common and often misdiagnosed as type 2 diabetes in clinical practice, new data show. 

The findings were presented October 4 here at the European Association for the Study of Diabetes (EASD) 2018 Annual Meeting by Nicholas J. Thomas, MD of the University of Exeter, United Kingdom.

In previous research on data from the UK Biobank, the Exeter team used genetic risk scores to show that type 1 diabetes — formerly called juvenile diabetes — is nearly as likely to occur after the age of 30 years as it is in younger people.

Those data were first presented at the 2016 EASD meeting and subsequently published in Lancet Diabetes & Endocrinology in 2017.

Now, the Exeter team has expanded that work to identify common characteristics of such patients and find out what actually happens to them in clinical practice. As hypothesized, in the current study of nearly 600 people with insulin-treated diabetes diagnosed after age 30 years, a large proportion who developed type 1 diabetes after age 30 years had been first diagnosed as having type 2 diabetes and were not treated with insulin initially, Thomas reported.

"In older ages, there's a dramatic increase in type 2 diabetes. Less than 5% have type 1 diabetes. So, you're trying to pick out a needle in a haystack," Thomas told Medscape Medical News, adding that because the prevalence of overweight and obesity is as high among people with type 1 diabetes as in the general population, "you're trying to find the minority where they often look very similar. It's really, really hard."

But one key sign identified from this study is rapid deterioration in glycemic control despite use of oral glucose-lowering medications. "Rapid insulin requirement should alert the clinician to the possibility of type 1 diabetes," Thomas said.

The correct diagnosis is important, he said, because patients need insulin early in order to avoid diabetic ketoacidosis. Also, even once they do receive insulin treatment, it's important to correctly distinguish the diabetes types as the respective insulin regimens and the type of education the patient receives differ considerably. Also, in the UK, patients with type 2 diabetes aren't eligible for supplies such as insulin pumps or continuous glucose monitors, whereas those with type 1 diabetes often are, Thomas noted.

But session moderator Colin Dayan, MBBS, professor of clinical diabetes and metabolism, Cardiff University School of Medicine, is a bit more hesitant about putting people on insulin straightaway. "Insulin is a very difficult drug to use...I don't want to start someone on insulin if they don't need it," he told Medscape Medical News in an interview. 

The issue, Dayan said, "is whether any damage is done by waiting, using tablets...There are certainly patients where we suspect that they may have type 1 diabetes but they didn't present with ketoacidosis, so we might start them on [an oral diabetes medication] with close observation, unlike somebody where we're confident they've got type 2 diabetes...You need to keep your eyes open the first 3 years."

Rapid Insulin Requirement Predicted Severe Endogenous Deficiency

Study participants were adults with diabetes in South West England between 2007-2017 who were diagnosed in primary care practices after age 30 years, treated with insulin, and had C-peptide levels measured.

Type 1 diabetes, eventually identified in 21% (n = 123) of patients, was defined as continuous insulin treatment beginning within 3 years of diagnosis and severe insulin deficiency (C-peptide < 200 pmol/L).

Another 306 people were defined as having type 2 diabetes, based on C-peptide ≥ 600 pmol/L after diabetes duration of at least 3 years (to avoid the type 1 diabetes “honeymoon” phenomenon, during which some insulin secretion persists for a time after diagnosis).

Patients with C-peptide levels between 200 and 600 pmol/L were considered indeterminate and not included in the analysis (n = 115).

For comparison, another group of 220 participants who met the criteria for type 1 diabetes diagnosed at age 30 years or younger were also included.

Among the 123 patients diagnosed after age 30 years who had severe insulin deficiency, 38% had not received insulin treatment at diagnosis and 47% self-reported that they had type 2 diabetes. 

Rapid insulin requirement was highly predictive of late-onset type 1 diabetes, with 84% requiring insulin within 1 year. And of all the patients treated with insulin within 3 years, 57% developed severe endogenous insulin deficiency consistent with type 1 diabetes.    

Asked about the reported slower-onset latent autoimmune diabetes of adulthood, or LADA, phenomenon, Thomas replied, "We're not seeing evidence of that...84% progressed to insulin within a year. That's not a slow decline."

Compared to the group with younger-onset type 1 diabetes, those with older-onset type 1 diabetes were similar in terms of body mass index (BMI), autoantibody positivity, and genetic risk scores.

In contrast, they differed markedly from those with type 2 diabetes diagnosed after age 30 years in that they had lower average BMI (25.9 vs 31.6 kg/m2), were much more likely to have positive islet autoantibody tests (78% vs 6%), and had higher type 1 diabetes genetic risk scores (0.268 vs 0.229).

But the management of those with later-onset type 1 diabetes differed significantly from those diagnosed with type 1 diabetes at younger ages, in that they were much less likely to have received insulin at diagnosis (62% vs 96%).  

No Perfect Test Discerns Diabetes Types, but Use "Clinical Suspicion"

Both Thomas and Dayan noted that there is no perfect laboratory test to distinguish between the two types of diabetes, as C-peptide can initially be positive in type 1 diabetes — although it's more definitive after about 3 years — and autoantibodies can be present in some people with type 2 diabetes and absent in some people with type 1 diabetes.

But, Dayan said, "There is clinical suspicion. When you have clinical suspicion, you monitor people more closely. We've got ketone testing, and there are lots of things we can do to prevent [DKA] happening if you know there's a potential risk."

Thomas said that autoantibody testing may be helpful for patients who are already suspected of having type 1 diabetes for clinical reasons, but of course can't be used to screen everyone.

His team is now trying to develop algorithms that will help clinicians distinguish the types more accurately, noting "the aim of our research is to identify the problem, characterize these people, and to try and make it easier for clinicians to ultimately get the best care for patients."

The study was supported by the Wellcome Trust, Medical Research Council, and Diabetes UK. Thomas has received a National Institute for Health Research academic clinical fellowship. Dayan is an advisor for Novo Nordisk, Sanofi, Genzyme, Servier, and Lilly.

European Association for the Study of Diabetes (EASD) 2018 Annual Meeting; October 4, 2018; Berlin, Germany. Abstract 301.

For more diabetes and endocrinology news, follow us on Twitter and on Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.