5 Subgroups of Diabetes You Should Know About

A game-changer in classifying the disease?

The American Diabetes Association classifies diabetes into four major categories: type 1 diabetes (insulin deficient), type 2 diabetes, gestational diabetes (diabetes in pregnancy), and specific types of diabetes due to other causes, such as maturity-onset diabetes of the young, disease of exocrine pancreas, latent autoimmune diabetes in adults.

Recently, in a study published in The Lancet Diabetes and Endocrinology, Swedish researchers characterized five subgroups of diabetes, each varying in severity from mild to severe. While the classification of type 1 diabetes remained unchanged, they grouped type 2 diabetes into four distinct subgroups.They believe that classifying diabetes using these subgroups helps to identify people with diabetes at most risk of developing complications

In addition, researchers believe this categorization will provide physicians with some therapeutic guidance—moving them one step closer to providing a personalized medication regimen. Optimally treating patients from the onset, based on the severity of their condition, could reduce the risk of diabetes-related complications.

How Subgroups Help

There is no one-size-fits-all when it comes to treating diabetes. Most people won't benefit from the same diet plan to lose weight and regulate blood sugars, similar to how most people with diabetes won't benefit from the same medication regimen.

While both the American Diabetes Association (ADA) and the American Association of Clinical Endocrinologists (AACE) have specific algorithms that they suggest clinicians use when prescribing medication, the ADA guidelines state that prescriptions should be based on a patient-centered approach, taking into consideration blood sugars, past medical history, age, efficacy, cost, potential side effects, effects on weight, hypoglycemia risk, and patient preferences.

The problem is that existing treatment strategies have failed at stopping the progressive course of diabetes. Swedish researchers believe that this shortcoming is because the diagnosis of diabetes is typically based on glucose alone. For example, two people presenting with elevated glucose can have varying degrees of disease progression, such as beta-cell loss (the cells that make insulin) and presentation. Hence, measuring glucose alone probably doesn't give us enough information to treat the disease as aggressively as we can.

Diagnosing patients using different subgroups of type 2 diabetes, based on characteristics such as insulin resistance and beta-cell dysfunction, can help classify the severity of their diabetes. As a result, medical strategies to best suit the individual patient can be developed—potentially making treatment more generalized in the future.  

Categorizing Subgroups

Researchers from Lund University in Sweden used data from the Swedish All New Diabetics in Scania cohort to do data-driven, cluster analysis in 8,980 patients with newly diagnosed diabetes. Subgroup classification was validated in three independent cohorts.

They grouped participants based on six variables, including the presence of glutamate decarboxylase antibodies (GADA, used to differentiate between type 1 and type 2 diabetes), age at diagnosis, body mass index, HbA1c level (a three-month average of blood sugar), and estimates of beta-cell function and insulin resistance. 

Using these variables, researchers were able to identify five different types of diabetes, some more severe than others. Severe autoimmune disease was the only subgroup of type 1 diabetes (which remains unchanged), whereas the other four were subgroups of type 2 diabetes.

Cluster Name Description Number/Percentage
Cluster 1 Severe autoimmune disease (SAID) early-onset disease, relatively low body mass idex (BMI), poor metabolic control, insulin deficiency, and presence of GADA 577 (6.4%)
Cluster 2 Severe insulin-deficient diabetes (SIDD) No presence of GADA, relatively low BMI, low insulin secretion, and poor metabolic control 1575 (17.5%)
Cluster 3 Severe insulin-resistant diabetes (SIRD) Insulin resistance and high BMI 1373 (15.3%)
Cluster 4 Mild obesity-related disease Presence of obesity, but no insulin resistance 1942 (21.6%)
Cluster 5 Mild age-related diabetes (MARD) Older patients than other clusters, similar description to cluster 4, but only modest metabolic derangements 3513 (39.1%)


The researchers compared disease progression, treatment, and development of diabetes-related complications between clusters. They found that people who were in clusters 1 and 2 had substantially higher hemoglobin A1c's at diagnosis than other clusters. Ketoacidosis at diagnosis was more common in cluster 1, which makes sense, since this cluster presents with insulin deficiency and presence of GADA (two determinants of type 1 diabetes). Cluster 3 had the highest prevalence of nonalcoholic fatty liver disease.

They also found that those with more severe forms, such as those who were severely insulin-resistant (cluster 3), had a significantly higher risk of developing diabetic kidney disease compared to other groups. Additionally, retinopathy (diabetes related eye disease) was higher in those who were severely insulin deficient (cluster 2). Cluster 5, made up of older patients with type 2 diabetes, had the most benign disease course. 

During their study, they found that treatment did not correspond with the type of diabetes.

Limitations to the Study

Unfortunately, we can't generalize this information to a global population because the study was derived mainly from Scandinavian patients. Future studies will need to look at more varied populations. 

In addition, we cannot determine if a person's classification changes as they age. We know that diabetes is a progressive disease—the longer a person has it, the more likely they will need intensive treatment (such as insulin) because, as the disease progresses, the beta cells that make insulin can become sluggish and die.

Researchers also only measured two types of autoantibodies. Testing additional autoantibodies can give us more information about the types and staging of diabetes. They also did not take into account additional risk factors for complications, such as blood lipids, triglycerides, LDL, HDL, cholesterol, blood pressure, and smoking. 

The ability to put this system of classification into clinical practice is unlikely for several reasons. First, researchers measured c-peptide concentration, which is not always measured in the clinic unless it is determined that it's needed to make a differential diagnosis. In addition, measuring insulin resistance and beta-cell function is not common practice.

Lastly, researchers used a sophisticated software program to determine clusters. This is not something that can be done in common-day practice, although they did suggest that a web-based tool to assign patients to specific clusters is under development. We will have to see how this unfolds. 

What Does This Mean?

Underlying physical process of disease continues to be more developed in type 1 diabetes than in type 2. Research conducted using first-degree relatives of patients with type 1 diabetes who persistently present with two or more autoantibodies is almost a certain predictor of hyperglycemia (high blood sugar) and diabetes. The progression is dependent on the age of detection of the antibody, the amount of antibodies, the type of antibody, and titer. The ADA has developed three distinct stages of type 1 diabetes which can help future research and treatment well before the onset of diabetic ketoacidosis. 

The issue still remains, though, of how to determine and understand beta-cell loss and dysfunction in people with type 2 diabetes. This path continues to be very unclear and varies widely from person to person. The one thing that continues to be a common denominator is that people who are insulin-resistant also have reduced beta-cell secretion. A system that enables us to classify type 2 diabetes into varying severities could be very helpful. In fact, the ADA, states, "future classification schemes for diabetes will likely focus on the pathophysiology of the underlying β-cell dysfunction and the stage of disease as indicated by glucose status (normal, impaired, or diabetes)." Being able to predict the progression of the disease will be critical in effective treatment strategies, but unfortunately we are just not there yet. 

What Happens Next?

We can use this classification system as a proposal, which, at the current moment, offers little clinical value for studying people with type 2 diabetes. Perhaps researchers will adopt these principals of categorization in the future, and perhaps they won't. But this study, at the very least, demonstrates the need to treat each patient with diabetes as an individual. For example, for those patients who appear to be insulin-resistant and deficient, perhaps intensive insulin therapy can help to stabilize blood sugars and prevent complications. Unfortunately, this hasn't seemed to be advantageous for children, but maybe it will work in adults.

Using this study as a platform may also help physicians think differently about their patients, thereby improving their treatment plans. 

Lastly, we should also be on the lookout—while the study was mostly limited to Scandinavian involvement, similar studies are in the works in China and India. It would be interesting to see if the results are the same. If so, perhaps a more generalized classification system can be implemented. 

A Word From Verywell

Years of evidence suggest that early and effective treatment for diabetes is critical in preventing life-shortening complications. More patient-specific treatment strategies may enable us to slow down the progression of diabetes and reduce the risk of diabetes complications. Dividing people with type 2 diabetes into subcategories varying in severity can help us craft optimal treatment strategies. Unfortunately, this classification system cannot be created yet, but it is becoming more clear that the more we know about diabetes progression, the better we can serve the people living with it. In the mean time, people with diabetes should continue to work on lifestyle management, such as healthy eating and exercise, which can only benefit them.

Was this page helpful?
Article Sources
Verywell Health uses only high-quality sources, including peer-reviewed studies, to support the facts within our articles. Read our editorial process to learn more about how we fact-check and keep our content accurate, reliable, and trustworthy.
  1. Ahlqvist, E., Storm, P., Käräjämäki, A. et al. Novel subgroups of adult-onset diabetes and their association with outcomes: A data-driven cluster analysis of six variables. Lancet Diabetes Endocrinol 2018. DOI:10.1016/S2213-8587(18)30051-2.

  2. Buse JB, Wexler DJ, Tsapas A, et al. 2019 Update to: Management of Hyperglycemia in Type 2 Diabetes, 2018. A Consensus Report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetes Care. 2020;43(2):487‐493. doi:10.2337/dci19-0066

Additional Reading