Short summary

We wish to quantify and describe how diabetes progresses in an entire population from before the actual diagnosis is established and through the many possible paths diabetics can follow in the course of the disease. We will mainly be looking at this disease progression in terms of the complication levels associ-ated to specific physiological domains (CNS, Heart, Kidney, Peripheral Nerves and Eyes) as well as the pace with which the transitions take place.

Rationale

Diabetes is a complex chronic disease with potentially serious consequences for those inflicted by it as well as great financial costs to the health care system for lifelong treatment. Etiologically, diabetes has traditionally been categorized into Type 1 diabetes and Type 2 diabetes as the two main types, reflecting the pathological mechanisms of the failure of producing insulin and the inadequacy of insulin to perform its crucial tasks in the body, respectively. This has served as a useful diagnostic framework that captures the most general characteristics of the disease. However, In recent years there has been a growing understanding that diabetes is much more complex and that a high degree of inter-patient variability exists. Even though diabetes, regardless of type, for most people will lead to cumulative organ damage with increasing malfunction over time, it is hard to predict which functions will be affected, in which order and the speed of progression of complications. Some diabetic patients are diagnosed early in life, others late, some progress quickly, while others are largely complication free. A number of risk factors, both genetic and lifestyle related, have been identified, but there is great variability in terms of the conse-quences and seriousness of the presence of these factors in the individual patient. The development of statistical progression models, classification algorithms and the identification of distinct progression typol-ogies may lead to a better understanding of the progression of diabetes and, as a consequence, lead to earlier and more targeted treatment plans for the individual.

By utilizing the unique population wide register data for all known Danish diabetic patients we hope to make new advances in this area.

We describe a diabetic state for a patient at a given point in time by a number of dimensions reflecting the complication state of central physiological domains relevant to diabetes. Initially, those dimensions are heart, kidney, CNS, eyes and peripheral nerves. For each such dimension we allow 3 states: unaf-fected, mild complications and severe complications. Whether the patient is diagnosed with diabetes or not, at a given point in time, is a separate binary dimension and it is possible for a dimension to be af-fected while a diabetes diagnosis is not yet established. A complication status for a given dimension does not necessarily have to be caused by diabetes, but rather reflects a disease burden at that point in time. The basis for assigning state values to dimensions, at any point in time, to any diabetic in the corpus, is an annotation of all diagnosis and procedure codes appearing in the LPR register. Each code has been annotated as diagnostic or non-diagnostic for diabetes and further annotated for its ability to assign a complication state to a specific dimension. Many codes are both diagnostic and indicative of complications at the same time while others will update only one of them. Based on the occurrence of medical events in the patient history, that are associated with annotated codes, we can describe and represent trajecto-ries as a series of state transitions and the number of days between them, that a patient goes through.

Description of the cohort

The net population for the study will be all Danish citizens registered in the national diabetes register (NDR) as of November 2014, approximately 500.000 individuals.

Data and biological material

For the above mentioned diabetes population, we have full history from the national patient register (Landspatientregisteret - LPR) going back to 1977 and data from the national diabetes register. This will include information about diagnosis, procedures and administrative codes pertaining to all hospital con-tacts of the individual. We also have CPR register about vital status of each individual.

Collaborating researchers and departments

Translational Disease Systems Biology Group - Novo Nordisk Center for Protein Research - University of Copenhagen

• Christian Simon, PhD-student
• Professor, Research Director Søren Brunak