Published on 8 March 2018

Swedish study proposes a new classification system for different types of diabetes.

A new classification system for different types of diabetes has been unveiled by researchers in Sweden.

The system replaces the current diagnosis of either type 1 or type 2 diabetes and proposes five different sub-categories, which could help healthcare professionals predict risk of serious complications and improve treatment suggestions.

The big difference from the current accepted classification of the condition is that type 2 diabetes consists of several sub-groups. 

The ANDIS (All New Diabetics in Scania) looked at people newly diagnosed with diabetes in Scania, southern Sweden and the findings from the study were recently published in The Lancet Diabetes & Endocrinology.


Headshot of Professor Leif Groop.

Professor Leif Groop

Leif Groop, Professor of Diabetes and Endocrinology at Lund University, study author and retired Chair of the DRWF Research Advisory Board in Sweden, said: “This is the first step towards personalised treatment in diabetes.

“Today's diagnosis and classification of diabetes are insufficient and can not predict future complications or choice of treatment.”

Professor Groop added that he believed the results represent a paradigm shift in how to look at the condition in the future. 

Professor Groop said: “Today, diagnosis is made by measuring blood sugar. If you also take into account the factors we have done in the ANDIS study you can provide a more accurate diagnosis.”

Diabetes is currently divided into type 1 diabetes (around 10% of people with diabetes), type 2 diabetes (85-90%) and a number of less common diseases such as LADA (latent autoimmune diabetes of adults), MODY (maturity onset diabetes of the young) and secondary diabetes.

The ANDIS study suggests the following classifications for diabetes diagnosis:

  • Group 1, SAID (severe autoimmune diabetes): In principle, this category corresponds to a diagnosis of type 1 diabetes and LADA (latent autoimmune diabetes in the adult), characterised by low-age disease, poor metabolic control, impaired insulin production and the presence of GADA (glutamic acid decarboxylase) antibodies - the most important factor for prediction of insulin therapy.
  • Group 2, SIDD (severe insulin-deficient diabetes): includes people without high HbA1C antibodies, impaired insulin production and moderate insulin resistance. Group 2 has the highest incidence of retinopathy.
  • Group 3, SIRD (severe insulin-resistant diabetes): characterised by severe obesity and severe insulin resistance. Group 3 had the highest incidence of kidney damage.
  • Group 4, MOD (mild obesity-related diabetes): includes people who are very overweight from a relatively young age.
  • Group 5, MARD (mild age-related diabetes): is the largest group (approximately 40%) and collects the oldest patients.

Professor Groop explained: “Those who benefit most from the new diagnostics are the most insulin resistant (Group 3) because they are currently the most misguided.”

For the study researchers looked at 13,720 people, aged 18 to 97, who were diagnosed with diabetes from 2008. By combining measurements of insulin resistance, insulin production, blood sugar levels and age at diagnosis, researchers found they could distinguish five distinct sub-categories that differ from today's classification of type 1 and type 2 diabetes.

Researchers have since repeated the study in another three studies from Sweden and Finland. 

Professor Groop said: “The outcome exceeded our expectations and was very well in line with the analysis from ANDIS. The only difference was that Group 5 was larger in Finland than in Sweden. The development of disease was remarkably similar in both groups.”

The recruitment of newly diagnosed diabetes patients continues and the researchers have several studies in turn based on the data already available. 

In addition to a more refined method of classification for people newly diagnosed with diabetes, researchers could also see that different groups had different risk for various health complications.

Emma Ahlqvist, Assistant Professor at Lund University Diabetes Centre and study author, said: “It means we can provide a treatment earlier that could prevent complications in those people with diabetes at greatest risk of suffering.

“The longer the study is going on, the more and better data we get, which gives us even better opportunities to tailor the treatment to each individual.”

The researchers plan to launch similar studies in China and India with people of other ethnic backgrounds.

It is estimated that around 425 million people in the world have diabetes. By 2045, the number is expected to have increased to 629 million.

Dr Eleanor Kennedy, DRWF Research Manager, said: “This is fascinating research which clarifies what many physicians have been seeing in their clinics for years – there are definitely more than the two sub-types of diabetes that we’ve been used to – type 1 and type 2 – for so long.

“It is challenging after all these years to think of diabetes now being referred to as “mild,” however, and I hope that people will not misconstrue the disease as not being important because of this.

“All forms of diabetes are serious and should be treated accordingly and it will be of interest to see what other researchers follow up on this work to explore different cohorts of people with diabetes in other parts of the world to see if they concur with the findings in Scandinavia.”

Read the report in The Lancet
Support DRWF by making a donation here
Find out more about DRWF-funded research here
Find out more about DRWF fundraising here
For latest update follow DRWF on FacebookInstagram and Twitter
To receive the charity’s latest bulletins as they become available, please sign up here
Read DRWF diabetes information leaflets here
Join the Diabetes Wellness Network here

Recent News