“Many patients got in touch after our first study, where we presented our classification of diabetes into five subgroups. They didn’t recognise themselves in the common idea of what a type 2 diabetes patient looks like and were grateful that we shed light on how much diabetes and the disease progression can differ between people. Of course, many physicians who treat patients had also noticed that diabetes is a heterogeneous disease, which our study could confirm," says Emma Ahlqvist, diabetes researcher at Lund University Diabetes Centre who leads research on subgroups of diabetes.
Emma Ahlqvist and her colleagues published the study in The Lancet Diabetes & Endocrinology in 2018, which showed that it is possible to divide people with type 2 diabetes into four subgroups and people with type 1 diabetes and LADA (latent autoimmune diabetes in adults) into one group, using a clustering method. Other researchers have since used the same method to replicate the classification system in over 20 cohorts in various populations.
The results hold
Emma Ahlqvist's research group now presents a new study on the subgroups in The Lancet Diabetes & Endocrinology. In the new study, the researchers have investigated whether the clustering approach is applicable on a new group of participants in the population study ANDIS (All New Diabetics in Skåne) and how the disease progresses in previous participants. The researchers have followed participants from the time they were diagnosed with up to 14 years of follow-up.
“Replication is key to verify and validate results, which we have done in our data-driven analysis. The previous results hold well in our new study. In this study, we have also increased follow-up time, the number of participants, and looked at data before diagnosis," says Olof Asplund, diabetes researcher at Lund University who conducted the statistical analysis.
The first article was based on data from 8,980 participants in ANDIS, while the new paper includes 19,076 participants with diabetes in the same population study. The researchers have confirmed that the classification can be applied to the new participants. The research team is also presenting several new findings of significant differences between subgroups when it comes to the development of diabetic complications and comorbidities. This part of the analysis has been carried out on both new and former participants within ANDIS.
High-risk group
One of the groups that stands out in the new study is SIRD (Severe insulin-resistant diabetes). This group is characterised by late onset of type 2 diabetes and obesity, severe insulin resistance and an increased risk of non-alcoholic fatty liver disease and nephropathy, also called diabetic kidney disease. The new study shows that SIRD is a high-risk group when it comes to the development of complications.
“This group would probably benefit from earlier identification and treatment, since this is a group with many risk factors for diabetic kidney disease and cardiovascular disease. Individuals in this group can probably have insulin resistance for a long time before being diagnosed with diabetes. Insulin resistance can cause comorbidity, and this means that a lot of damage can occur before individuals in this group are diagnosed with diabetes," says Emma Ahlqvist.
People in SIRD share several complications with the SIDD group. Both groups had an increased risk of kidney disease, myocardial infarction, and mortality. The MOD group has previously been described as a group with a relatively mild disease progression, but the new study showed that this group also had an increased risk of mortality.
“People in the MOD, SIDD and SIRD groups had an increased risk of mortality, largely driven by cardiovascular disease, even after we had adjusted for established risk factors. People in MOD have early onset of type 2 diabetes and will live with the disease for many years. Given the increased risk of complications, it cannot be regarded as a mild form of diabetes. We will therefore change the classification of this form of diabetes from mild to moderate," says Emma Ahlqvist.
Knowledge that can reach the clinic
More evidence is needed to offer individualised treatment based on the classification of diabetes in the clinic. The research shows that there is a small group of people with diabetes who cannot be assigned to any of subgroups, which may need to be studied in more detail. The researchers hope that the knowledge about the subgroups can be implemented in the clinic in the future, to assess disease risk, resource allocation, and provide individualised treatment of patients. The researchers are now investigating how different groups of patients respond to different kinds of treatments.
“We have started to analyse data on blood sugar levels in relation to the drugs different groups take, and how many complications they develop during treatment. If certain drugs result in better blood sugar control and fewer complications in certain groups, it may be interesting to follow up the results in controlled treatment studies," says Olof Asplund.
