CHICAGO – “Lean type 2 diabetes (T2D)” is a diagnosis of exclusion but is also a true entity affecting roughly 10%-20% of people with T2D.
That was the message from Kristina Utzschneider, MD, associate professor of medicine at the University of Washington, Seattle, and director of the Diabetes Care Program at the Veterans Affairs Puget Sound, at the American Diabetes Association (ADA) 85th Scientific Sessions in a symposium about “nuanced” treatment of T2D.
“Type 2 diabetes is a condition characterized by hyperglycemia due to insufficient insulin secretion typically occurring in the context of insulin resistance, after ruling out other potential causes of diabetes. I think that last phrase is key,” she said. “The first thing is to consider the differential diagnosis, especially if you’re seeing a lean patient, which is a little atypical. Could diabetes be due to something else?”
Two of the most common alternative diagnoses are autoimmune (type 1 diabetes [T1D]) diabetes or a monogenic form called MODY. The former can be assessed by measuring autoantibodies (GAD, IA-2, or ZnT8), while the latter is characterized by onset younger than 35 years, a strong family history of diabetes, and negative autoantibodies. Other less common alternatives include the result of pancreatic damage, cystic fibrosis-related mitochondrial diabetes, Wolfram syndrome, or lipodystrophy, Utzschneider said.
“In my clinical practice, with most of the lean patients who have come to me because they’re not doing well, I test them, and they’re type 1,” she told Medscape Medical News.
However, she added, a sizable minority of people who actually do have T2D are lean, pointing to the heterogeneity of the condition. “Lean type 2 diabetes is not necessarily rare. Beta cell dysfunction and insulin resistance are key factors underlying T2D, with more of a beta cell insulin secretory defect in the lean T2D group, but some patients may still have significant insulin resistance that could be targeted.”
Very little research has focused on that group, she noted. “It’s tough finding data, as most of the population with type 2 is not lean. In most of the big studies, especially about treatment, the average BMI is 33.”
Also in the mix are people of Southeast Asian ethnicity, who typically develop T2D at a lower average BMI than people of other ethnicities.
Session moderator Liana K. Billings, MD, director of the Diabetes and Obesity Research Program and the Personalized Medicine in Diabetes Program at NorthShore University HealthSystem, Skokie, Illinois, endorsed the recommendation that for any lean person diagnosed with diabetes, the first step is to check whether they have type 1 or a monogenic type.
“Those tests are becoming more and more available, like the type 1 polygenic risk scores that can be used in helping to understand the genetic risk someone has for type 1 versus type 2 diabetes. It’s definitely a growing field where we’re trying to make sure that people with lean diabetes get the right diagnosis right out of the gate,” Billings said.
“We’re trying to understand how people with lean type 2 may respond differently to medications, or what their prognosis is, and what factors may increase a person’s risk of having diabetes at a lean BMI vs a higher BMI,” she added.
Lean and Insulin-Resistant
According to the World Health Organization, classifications of “lean,” “overweight,” and “obesity” differ in people of Asian background from those of other ethnicities, with cutoffs of < 23, 23-27.4, and ≥ 27.5, respectively, vs < 25, 25-29.9, and ≥ 30 in people of other ethnicities.
“Lean does not necessarily mean someone is insulin sensitive, although typically as BMI increases, insulin sensitivity decreases. A lean individual could potentially be somewhat insulin resistant. And vice versa, somebody who is obese is not necessarily insulin-resistant,” Utzschneider noted.
According to data from the US CDC for 2021, approximately 10.2% of people with diagnosed diabetes have a BMI < 25, although that includes people with T1D. However, in contrast, Korean data for 2019-2020 show the proportion of people with diabetes who have a BMI of 18.5-22.9 is about 23.3%. “Think about this when you’re seeing patients in your clinic. If they have Asian ancestry, particularly South Asian ancestry, your threshold for considering [type 2 diabetes] is going to be higher.”
Different Treatment by BMI? More Research Needed
Very little data are available regarding the effects of T2D treatment in lean people specifically. However, one study evaluated 20 adults with T2D for < 6 years and BMI < 29 not on insulin who were put on a very low-calorie diet aiming for a 5% weight loss, then returned to a regular diet. Their BMIs and visceral and liver fat all declined. Glycemic control and fasting glucose declined as well, and 70% achieved T2D remission. Insulin sensitivity doubled, and beta cell function improved from baseline but did not reach the level of control. “They’re managing the diabetes with the weight loss and lifestyle, but the beta cell defect is still there,” Utzschneider said.
As for medications, most studies haven’t shown a huge effect of BMI on A1c-lowering effectiveness, with a few exceptions. In the ADOPT study, rosiglitazone was more effective at lowering A1c than metformin among those with obesity. And in a precision medicine modeling study, DPP-4 inhibitors were less effective in people with BMI ≥ 30 or who were very insulin-resistant.
In general, Utzschneider said, metformin is still the first-line option, and other glucose-lowering medications are worth trying, but those with lean T2D may require insulin sooner than those with higher BMIs. This can be assessed by measuring C-peptide, which reflects endogenous insulin secretion. It’s important to measure both glucose and C-peptide at the same time because if glucose is low, “the body will shut off insulin secretion, and you won’t be able to interpret the results,” she noted.
There are no clinical practice guidelines for checking C-peptide in T2D, and no set cutoffs. Utzschneider considers a value of < 0.6 ng/mL to be low, likely requiring insulin, and a value of > 2.7 ng/mL to be high, indicating that the individual is still making insulin but is likely very insulin-resistant, so the treatment should focus on that.
Caveats about C-peptide testing include the fact that the C-peptide assay isn’t standardized so there’s variability from lab to lab, it can be affected by renal clearance and certain medications, and the measure won’t be accurate in the setting of glucotoxicity. “This is an area we need more research and better guidelines on,” she said.
Ongoing Research Is Digging Into T2D Heterogeneity
Better subclassification of T2D and predictive models are also needed. In one recently published study, researchers from Exeter University, Exeter, England, devised a model for A1c lowering at 1 year, based on nine readily available clinical variables including baseline BMI to recommend the optimal drug class for that patient. “This is really exciting work,” said Utzschneider, who added that her team is collaborating with the Exeter group on further analyses.
She is also participating in a research consortium funded by the National Institute of Diabetes and Digestive and Kidney Diseases called DEFINE-T2D with the goal of improving the classification of T2D by integrating multiple data types from large existing datasets and ultimately leading to improved treatment outcomes.
Utzschneider reported receiving research funding from Eli Lilly and Company, Avid, and Amgen and consulting fees from Nevro Corp. Billings reported receiving research support or consultant fees or served on advisory panels for Bayer, Dexcom, Endogenex, Eli Lilly and Company, Novo Nordisk, Sanofi, Pfizer, and Xeris.
Miriam E. Tucker is a freelance journalist based in the Washington, DC, area. She is a regular contributor to Medscape Medical News, with other work appearing in the Washington Post, NPR’s Shots blog, and Diatribe. She is on X @MiriamETucker and BlueSky @miriametucker.bsky.social.