The treatment of Type 1 and Type 2 diabetes is very different: People with Type 1 diabetes rapidly stop making their own insulin, so need insulin injections from diagnosis. People with Type 2 diabetes can keep making their own insulin but it may not work as well as it should, so they can be treated with diet or tablets. While they may eventually need insulin treatment it is usually not until many years after diagnosis.
It is often difficult for doctors to tell which kind of diabetes a person has. Because of this, sometimes people are given the wrong diagnosis. This can have a huge impact as it means they could receive the wrong treatment.
This study aims to improve this situation by helping doctors more accurately tell the type of diabetes a person has when they are first diagnosed.
We will recruit 1000 participants who have recently been diagnosed with diabetes between the ages of 18 and 50. We will recruit an additional 800 recently diagnosed participants who developed diabetes after age 50, half of whom will be receiving insulin therapy. We will record clinical features and measure blood tests that may help us determine diabetes type at diagnosis and follow participants for 3 years to see whether they stop producing their own insulin and need insulin treatment, which confirms Type 1 diabetes. We will assess whether clinical features and blood tests can help us tell if a patient needs rapid insulin treatment and should be initially treated as Type 1 or Type 2 diabetes.
Findings will be integrated into a clinical prediction model that will be freely available as a website calculator and smartphone app.
Primary objective
- To establish diagnostic performance of biomarkers including islet autoantibodies, C-peptide and a genetic risk score in identifying patients with rapid requirement for insulin, alone and in combination with clinical features.
Secondary objectives
- To prospectively validate a clinical prediction model developed from cross sectional datasets in predicting rapid insulin requirement in young onset diabetes.
- To integrate discriminative and additive biomarkers into the clinical prediction model.
- To establish a bio resource for future biomarkers discovery and assessment.
Primary endpoint/outcome
- Diabetes type defined by insulin requirement at 3 years:
- Type 1 diabetes = Progression to insulin treatment and severe insulin deficiency (post meal plasma C-peptide <0.6nmol/L) at 3 years
- Type 2 diabetes = Lack of requirement for insulin at 3 years (HbA1c <90mmol/molwithout insulin treatment or post meal C-peptide ≥ 0.6nmol/L if insulin treated)
Secondary endpoints/outcomes
- Stimulated C-peptide <0.2nmol/L at 3 years (‘absolute insulin deficiency’)
- C-peptide rate of change (UCPCR and plasma)
- HbA1c (mean and at 3 years)
- Weight change (baseline to 3 years)
- Self-reported hypoglycaemia & hypoglycaemic awareness
- Wellbeing and resilience (SF12 and CD-RISC questionnaire)
- Ketoacidosis (self-reported and confirmed from medical notes)
- This research will produce a robustly validated clinical prediction model that will for the first time integrate validated clinical features and biomarkers to allow clinicians to accurately assess classification and need for insulin treatment at diabetes diagnosis. The model will be freely available to clinicians and patients through a website and smartphone App.
