The onset of type 1 diabetes results from cumulative loss of beta cell mass and function. The main markers of this process are decline of the first-phase insulin response (FPIR) to injected glucose, together with slow deterioration in oral glucose tolerance. Pulsatile insulin secretion is also lost at an early stage but is much less easy to measure. The point at which a declining beta cell mass translates into glucose intolerance is mediated by the level of insulin resistance. Intervention trials have typically used both oral and intravenous glucose challenges as an indication of risk of imminent progression to type 1 diabetes, but the tests are cumbersome to perform and require careful standardisation.
Measuring beta cell mass and function
The beta cell is the target organ of type 1 diabetes, and progression to diabetes might therefore be most satisfactorily measured in terms of declining beta cell mass and/or function. The existence of a functional component is implied in the observation that partial recovery of insulin secretion may occur in the so-called remission or honeymoon phase following diagnosis. Beta cell mass is unfortunately impossible to measure with any precision by imaging techniques or tissue biopsy, and tests of insulin secretory function are therefore needed.
C-peptide provides a simple measure of insulin secretion, and is usually measured in the circulation, although urine C-peptide might potentially offer a useful integrated measure of insulin secretion, and is under investigation in this respect. Fasting plasma C-peptide in a normoglycaemic individual reflects the prevailing level of insulin resistance, and does not provide a reliable indication of insulin secretory capacity. For this reason stimulation tests are often used, either following intravenous injection of glucagon (a potent insulin secretagogue), or following a mixed meal. Repeated measurement of stimulated C-peptide may provide an indication of declining beta cell function in a given individual, but has limited value in the prediction of diabetes
First-phase insulin response (FPIR)
In contrast to oral administration, glucose injection into the circulation produces a biphasic insulin response, with an immediate first phase lasting ~10 minutes followed by a delayed second phase which peaks after 2–3 hours. Loss of the FPIR can be induced by partial pancreatectomy in experimental animals, suggesting that this is a useful index of beta cell mass in humans. Prospective studies have shown that loss of FPIR is an acquired defect in the prodrome to type 1 diabetes and highly predictive of early clinical onset. Loss of FPIR is associated with loss of pulsatile insulin secretion, impaired hepatic sensitivity to insulin, and decreased hepatic clearance of insulin.
In the Diabetes Prevention Trial – Type 1 (DPT-1), about 60% of first degree relatives positive for both ICA and IAA and with FPIR below the first percentile progressed to diabetes within 5 years. In the European Nicotinamide Diabetes Intervention Trial (ENDIT), the overall 5-year risk in ICA-positive relatives with low FPIR was about 55%, but among relatives positive for five autoantibodies (ICA, GADA, IA-2A, IAA, IA-2βA) the risk associated with a low FPIR increased to >90%, and impaired glucose tolerance identified those with the highest risk of progression within the following year. Combining measures of insulin resistance and FPIR in islet autoantibody-positive relatives may further contribute to risk assessment, but this approach is of limited practical application.
Oral glucose tolerance
Fasting glucose levels do not deteriorate until a late stage in progression to type 1 diabetes, and are therefore of limited predictive value. The oral glucose tolerance test (OGTT) provides a familiar measure of glucose tolerance, but has a high coefficient of variation around the key 2-hour value used in interpretation of the test. Repeated OGTTs provide a useful but late-stage indication of imminent progression to diabetes.
Metabolic testing adds to the logistic challenges of an intervention trial, is time consuming, and requires careful standardisation. It is also unpopular among the trial participants. At present there are few practical alternatives, but more targeted use of antibody testing might provide one possibble alternative. Low FPIR was associated with autoantibody characteristics such as high titre ICA, the presence of IAA and multiple islet autoantibodies in children aged 1–5 years in the DIPP study, and in older relatives participating in ENDIT,2 indicating that some of the increased risk conferred by low FPIR can be identified by autoantibody measurement. It has been proposed that accurate type 1 diabetes risk stratification can be achieved on the basis of autoantibody titre, subclasses and/or epitopes alone, suggesting that FPIR may not be an essential criterion for recruitment for future trials.
^ Diabetes Prevention Trial--Type 1 Diabetes Study Group. N Engl J Med 2002;346:1685–91
^ Bingley PJ, Gale EA. Diabetologia 2006;49:881-90
^ Fourlanos S et al. Insulin resistance is a risk factor for progression to type 1 diabetes. Diabetologia 2004;47:1661–7
^ Keskinen P et al. First-phase insulin response in young healthy children at genetic and immunological risk for Type I diabetes. Diabetologia 2002;45:1639–48
^ Achenbach P et al. Stratification of type 1 diabetes risk on the basis of islet autoantibody characteristics. Diabetes 2004;53:384–92