“This study gives new insights into the development of the disease in children who are at high risk of developing the disease. We have gathered data from five prospective studies from several different countries, which is a great strength. Our research colleagues at IBM Research have the knowledge of machine learning and data visualisation that this project has required,” says Markus Lundgren, researcher in pediatric endocrinology at Lund University Diabetes Centre (LUDC) and one of the authors of the study, published in Nature Communications.

The study in Nature Communications includes data from 24.662 children, who have been followed for 15 years. Data are derived from five prospective studies of children in United States, Sweden, Germany and Finland. Researchers at Lund University Diabetes Centre have provided data from the study Diabetes prediction in Skåne, where children in southern Sweden have been screened for type 1 diabetes risk from infancy up to 15 years of age.

Three pathways

The research team studied the pathway from being a healthy individual to progression to type 1 diabetes in 652 individuals. Until now, type 1 diabetes has often been divided into different stages depending on how many diabetes related autoantibodies the individual has got. The new study shows that type 1 diabetes instead can be divided into three different groups based on the patterns of developing autoantibodies.

“We have been able to identify three distinct trajectories, or pathways, from healthy individual to diagnosis, that are associated with varying degree of disease risk,” says Bum Chul Kwon, first author of the study and researcher of data visualisation and visual analytics at IBM Research.

In discussing this new study in Nature Communications, Markus Lundgren, principal investigator of the study in southern Sweden says:

“This may be seen as a first step towards a categorisation of patients with type 1 diabetes into subgroups. We have identified that the groups differ and need to find out more about how they differ by for example studying genetic factors that may play a role.” 

Screening for type 1 diabetes

It is currently not possible to prevent or cure type 1 diabetes, which is why research on prevention and early detection is important. The researchers behind the study hope that the new knowledge may pave the way for screening programmes for type 1 diabetes. 

“An important goal is to create a disease progression model that will be able to predict the disease with high accuracy. It is however of great importance that the health care system has more to offer people at risk of developing the disease before screening can be a viable option, such as a treatment that will delay the disease progression,” says Markus Lundgren.

The study has been funded by JDRF International. Frank Martin, senior director at the organisation’s research department, thinks that the study has transformative potential. 

“Only through better screening, monitoring, and risk education can we minimise the near- and long-term health consequences associated with missed diagnoses,” Frank Martin says.