Nearly 4% of people in the US initially diagnosed with type 2 diabetes (T2D) between July 2016 and October 2024 were later reclassified as having type 1 diabetes (T1D), a new analysis of data from the TriNetX database showed.
“We still have this problem of identifying people with type 1 diabetes,” said endocrinologist Jeremy H. Pettus, MD, associate professor of medicine at the University of California, San Diego (UCSD), who presented the study findings at the 19th International Conference on Advanced Technologies and Treatments for Diabetes (ATTD) 2026.
“We really have to think about this and move towards” using autoantibodies more often, he added.
Individuals who had been initially misdiagnosed tended to be younger than those who retained the T2D diagnosis, but there was no difference when stratified by BMI, Pettus reported. The reclassified group also had higher healthcare utilization.
150,000 People Misdiagnosed
In RECLASS-T1D, Pettus and colleagues retrospectively analyzed data from TriNetX electronic health records for a total of 6,759,145 people diagnosed with either prediabetes or T2D between July 1, 2016 and October 30, 2024. Overall, 2.2% (147,419 individuals) were later reclassified as having T1D. Among 3,400,462 initially diagnosed with T2D, 3.9% (129,866) were reclassified as T1D.
“And keep in mind, we’re only identifying the people who were reclassified, so there [were] lots of people, I imagine, who were still floating around with type 1 diabetes [who] never got the right diagnosis,” Pettus noted.
Of the total 6.7 million, the proportion who were reclassified didn’t differ by gender or race/ethnicity, but it did decrease with age, from 7.0% for those younger than 18 years, to 4.4% for ages 18-35, to 1.8% for those older than 35 years. (This was a descriptive analysis, so no statistical calculations were performed.)
In contrast to what might have been the case in the past, BMI was not a reliable predictor of reclassification. Proportions of patients who were reclassified ranged from 2.5% for those with BMI below 18.5 to 1.5% of both those with overweight (BMI 25-29) and obesity (> 30).
“I think clinically we feel that if a patient is on the heavier side they’re type 2 and if they’re leaner it’s type 1, but that’s really not the case at all,” said Pettus. “We can’t rely on BMI to readily distinguish if somebody has type 1 or type 2.”
Reclassification rates were higher among individuals with kidney disease (2.8%), but looking at their creatinine to tell if you should do antibody screening “isn’t something that I would rely on,” he noted.
Compared to those not reclassified, those who were had higher baseline A1c levels at diagnosis (8.3% vs 7.1%). Their mean age was lower but not young (49.8 vs 57.8 years), and their BMI was nearly identical, 31.1 vs 31.5.
Asked to comment on the results, session moderator Anastasia Albanese-O’Neill, PhD, APRN, CDCES, vice president for medical affairs at Breakthrough T1D (formerly JDRF), noted that beta cell loss in adult-onset T1D can progress more slowly than in childhood, which contributes to the confusion.
“But it’s still autoimmune, and that means we should consider intensive insulin therapy and think about clinical trials to preserve beta cells,” she told Medscape Medical News. “It’s a different diagnosis [from T2D], and there should be a different care plan.”
Broader Antibody Screening?
Individuals with the autoimmune conditions Addison disease and celiac disease also had higher reclassification rates (2.9% and 2.4%, respectively), noted Albanese-O’Neill, who was not involved in the study.
“When somebody’s not responding to therapy or they have a personal or family history of autoimmunity, clinicians should have a high suspicion for type 1 diabetes and order those autoantibodies,” she said.
“We need to make that clinical process, especially for primary care, more feasible. It cannot be that they have to download a form and fax the lab. The labs have to be bundled and easy to order for those folks in primary care who are our partners in this work,” she added.
Comparing those reclassified vs those who weren’t, the proportions having one or more diabetes-related healthcare-related resource utilization during the follow-up period (prior to T1D diagnosis among those reclassified) were 66.4% vs 38.5%. All types of encounters were greater among those reclassified, including any hospitalization (18.5% vs 11.6%), outpatient visits (42.4% vs 27.3%), and emergency department (13.8% vs 6.4%).
“Maybe they have [diabetic ketoacidosis], maybe they have early complications of diabetes. Getting the diagnosis wrong is both mentally disturbing to patients and also a big hit to the healthcare system,” said Pettus. “I’m not a healthcare economist, but this is the type of data I would imagine could start pushing us to say this is worth it in dollars and cents.”
Indeed, Pettus said that — at UCSD — the electronic health record now has a checkbox for autoantibodies rather than having to manually type in the names of specific tests. “I’m hoping we’re moving towards universal screening at diagnosis. With more screening in terms of identifying people before they have type 1, there’s a big push for using antibodies in general.”
Most reclassifications (81%) occurred within 3 years of the initial diagnosis, while 19% occurred beyond 3 years. Still, more than 4000 individuals weren’t correctly diagnosed for 7 or more years. “What about these people who aren’t picked up for 4, 5, 6 years? … I feel like this is a call to action,” said Albanese-O’Neill.
Pettus has served as an advisor to Sanofi, Novo Nordisk, and Kriya and on the speakers bureau for Sanofi. Albanese-O’Neill is employed by Breakthrough T1D.
No comments:
Post a Comment
Note: Only a member of this blog may post a comment.