Development of a new diagnostic tool for individualized treatment of diabetes
| Reference number | |
| Coordinator | Lunds universitet - Lund Univerity Diabetes Center |
| Funding from Vinnova | SEK 936 500 |
| Project duration | December 2014 - August 2015 |
| Status | Completed |
Purpose and goal
The aim of this project has been to create a diagnostic panel together with a newly formed company (Diatunnel) and Region Skåne . To this in end we have utilized data from a previously unpublished cohort of 10 700 newly diagnosed diabetics to perform a data driven identification of new diabetic subtypes as well as factors important for diabetic complications. Initial steps to patent this panel has also been taken.
Results and expected effects
The cluster analysis identified five subgroups of diabetes, which predicted change in HbA1c over time as well as development of early diabetic complications. All of the clusters had distinct genetic profiles and some showed changes in metabolites. A common variant in the SLC2A2 gene encoding GLUT2 was a strong predictor of time to insulin therapy and development of cardiovascular complications. The results were additionally made available through an web-application for prospective testing of the new subgroups.
Approach and implementation
The ANDIS project in Southern Sweden included 10,785 newly diagnosed patients aged 0-97 years. A similar registry from Western Finland, DIREVA (Diabetes Registry Vaasa), with 5,107 patients aged 0-94 years at diagnosis, was used for replication. We used information on six easily measured variables (age at onset, BMI, HbA1c, measures of insulin secretion and action, GADA autoantibodies) to identify factors predicting change in HbA1c over time, time to insulin therapy, development of early complications.