C-peptide in T1 prediction

The beta cell is the target of the immune process culminating in type 1 diabetes, which is associated with progressive loss of beta cell mass and function. Beta cell mass cannot be estimated directly, and functional changes in insulin secretion are therefore a useful surrogate. Both insulin and C-peptide are derived in equimolar amounts from their precursor, proinsulin, but C-peptide has the advantages that it is not cleared by the liver and can be measured more reliably. C-peptide secretion is around 20% of that of healthy volunteers at diagnosis of diabetes. The level of C-peptide is inversely proportional to age, as is the rate of decline in insulin secretion following diagnosis. Single fasting or post-prandial measures of C-peptide have limited value in diabetes prediction, and dynamic tests of insulin secretion such as the IVGTT are to be preferred. Loss of FPIR to below the first centile of a healthy population is highly predictive of diabetes onset within 2 years. Supplementary dynamic tests include the response to a mixed meal or to the OGTT. The gold standard test is the hyperglycemia clamp, which allows second as well as first phase insulin responses to be measured, but this is mainly useful for research purposes.

Biology of C-peptide

Figure 1
Figure 1
C-peptide and insulin both originate from the same precursor molecule proinsulin (Figure 1), and as a consequence they are secreted by the beta cells in equimolar amounts. Both hormones can be used as surrogate markers of beta cell function, but the use of C-peptide was recommended by an ADA workshop for several reasons[1]. These are that C-peptide offers a better reflection of pre-hepatic insulin secretion since – unlike insulin - it is not extracted by the liver but exclusively by the kidneys. Furthermore, it has a constant peripheral clearance, while clearance of insulin is subject to considerable day-to-day variation. Finally, C-peptide assays are often more reliable than insulin assays since many of these cannot differentiate insulin from proinsulin or conversion intermediates, and they are less precise in the lower concentration range typical of patients with type 1 diabetes.

Beta cell function around and after diabetes onset

Already in 1965 Willy Gepts [2]described residual beta cell mass shortly after diagnosis to be between 10 and 33% of that in non-diabetics. These data were confirmed by hyperglycemic clamp tests in new-onset patients, which found C-peptide secretion to be around 22% of that in healthy volunteers[3]. Residual C-peptide secretion at diagnosis is highly correlated with the age of the patient at diagnosis, being lowest in children diagnosed under age 7, whereas loss of beta cell function is less pronounced in patients above age 15.

A few weeks after diagnosis (and after starting insulin treatment) most younger patients have a decreased need for exogenous insulin together with increased endogenous insulin secretion. This phase of partial clinical remission, often termed the "honeymoon period", is present in half of the children diagnosed with type 1 diabetes and can last up to 24 months after diagnosis.

Residual functional beta cell mass declines further after diagnosis and ultimately disappears completely in the majority of patients. Most children diagnosed under age 7 will completely lose their insulin secreting capacity within 1 year. In contrast, 40% of the pancreata of patients diagnosed between age 7 and 30, still contained beta cells when autopsied 1-30 years later[1].

Decrease of insulin and C-peptide secretion before clinical onset

Beta cell function is therefore already reduced at diagnosis, and continues to decrease until it finally disappears in most patients. More subtle alterations in beta cell function become manifest months or even years prior to diabetes onset. In 1986 George Eisenbarth proposed a model assuming that beta cell function declines progressively after the onset of autoimmunity[4].

More recently several adaptations to this model were proposed. First, it was demonstrated that relatives who progress to diabetes do not show the rise in C-peptide secretion that is normally seen with increasing age. Consequently, the decrease in C-peptide secretion at diagnosis should not just be considered as a loss in absolute terms, but also as a failure to increase with age[5].

Second, it is hypothesized that in some patients the progression is not linear, but may occur in waves and cycles, and that type 1 diabetes could be a so-called relapsing–remitting disease[6]. With the progression of beta cell destruction there may be an increasing number of antigenic determinants, so-called epitope spreading, which leads to an increase in the immunologic response. Moreover, beta cells may not be innocent bystanders, but may contribute to the disease process via the mechanism of beta cell regeneration. This could be associated with the creation of new antigenic epitopes, reinforcing the immune response but producing temporary recovery of beta cell function at the same time.

The Diabetes Prevention Trial 1 (DPT-1 trial) yielded interesting information on the natural course of beta cell function in ICA-positive first-degree relatives of type 1 diabetic patients[7]. The mean C-peptide response during an oral glucose tolerance test (OGTT) declines prior to diagnosis. According to the DPT-1 data this decline in stimulated C-peptide levels starts at least 2 years before diagnosis and accelerates around 6 months prior to diagnosis. Fasting C-peptide values are maintained longer, even after diagnosis. The decrease in beta cell mass is thought to be more rapid in younger children, especially under age 7, while in adults the immune process seems to be less fulminant, resulting in better preserved beta cell function at diagnosis.

Residual beta cell function and prediction of type 1 diabetes

The measurement of functional beta cell mass before clinical onset is particularly valuable for the prediction of the disease in antibody-positive relatives, since autoantibodies only reflect the ongoing immune process and cannot be used as markers of residual beta cell function.

Estimation of beta cell function can complement immune and genetic markers to refine the prediction of diabetes, but fasting or random samples for C-peptide measurement are simply not sensitive enough and more laborious tests are needed. In 1984 Srikanta and co-workers described a progressive loss of first phase insulin release (FPIR) during an intravenous glucose tolerance test (IVGTT ) and showed for the first time that the sum of the insulin concentrations at 1 and 3 min has some predictive value [8].

Several subsequent studies showed a high risk for impending diabetes (between 75% and 92%, depending on the study) in ICA-positive relatives with a decreased FPIR. While most of these studies describe the association between a decreased FPIR during IVGTT and the risk for progression to diabetes, only a few described the association with the time to onset of the disease.

Vardi et al. noted that the absolute value of the FPIR was inversely associated with the time to diabetes onset and that a FPIR below the first percentile preceded the onset of diabetes by an average of 22 months[9]. The ICARUS study described an association between a low FPIR and the risk for rapid progression towards diabetes[10].

A more laborious but more reproducible test to measure beta cell function prior to diabetes onset is the hyperglycemic clamp test, described by de Fronzo in 1979. This test not only provides information on FPIR (representing the rapidly releasable pool of insulin in the islets of Langerhans) but also on insulin release following sustained stimulation with glucose, the so-called second phase insulin release (SPIR).

A hyperglycemic clamp study of high-risk IA-2A-positive first degree relatives showed that a low FPIR can identify those relatives who will progress rapidly towards clinical onset (positive predictive value 100% for progression within 24 months). This indicates that the loss of the FPIR may be a relatively late phenomenon in the preclinical phase. A decreased SPIR on the other hand (with preserved FPIR), was associated with slower progression to diabetes (positive predictive value 58%, progression after more than 34 months) and might represent an earlier stage of the disease.

Normal FPIR and SPIR seem not to be associated with progression towards clinical onset, despite the high-risk antibody profile of the relatives in this group[11]. More recent research has focused on the role of OGTT parameters in risk prediction, since all surveillance programs include OGTT testing for early diagnosis of diabetes. If measures obtained with the OGTT turn out to be equally predictive as the IVGTT, this would imply an important reduction of efforts and costs.

In conclusion

Prior to diabetes onset, there is a gradual loss of beta cell function, and it is not until beta cell function has decreased to about 20 % of the beta cell function in healthy volunteers that blood glucose level will rise and lead to the diagnosis of diabetes . The measurement of functional beta cell mass before clinical onset is particularly valuable for the prediction of the disease in antibody-positive relatives, since the autoantibodies only reflect the ongoing immune process and cannot be used as time-dependent markers of residual beta cell function.

The estimation of beta cell function can complement immune and genetic markers to refine the prediction of diabetes. C-peptide is a good measure of beta cell function. Ideally it should be determined during a dynamic test, since fasting or random samples lack sensitivity. Several tests have been described, but the IVGTT or mixed meal tolerance test are most often used. Decrease of the FPIR during IVGTT or hyperglycemic clamp test increases the risk for rapid progression to diabetes, but the second phase of the hyperglycemic clamp could potentially be used as a tool to identify relatives who will progress to diabetes at an even earlier stage. Use of OGTT parameters could be of particular interest since this test could combine measurement of beta cell function and early diagnosis of dysglycemia at the same time.

See also C-peptide in type 1 diabetes


  1. ^ Palmer J.P. et al. C-peptide is the appropriate outcome measure for type 1 diabetes clinical trials to preserve beta-cell function: report of an ADA workshop, 21-22 October 2001. Diabetes 2004; 53:250-264

  2. ^ Gepts W.: Pathologic anatomy of the pancreas in juvenile diabetes mellitus. Diabetes 1965; 14:619-633

  3. ^ Keymeulen B et al. Insulin needs after CD3-antibody therapy in new-onset type 1 diabetes. New England Journal of Medicine 2005; 352: 2598-2608

  4. ^ Eisenbarth G.S. Type I diabetes mellitus. A chronic autoimmune disease. N Engl J Med 1986; 314:1360-1368

  5. ^ Schatz D. et al. Preservation of C-peptide secretion in subjects at high risk of developing type 1 diabetes mellitus--a new surrogate measure of non-progression? Pediatr Diabetes 2004; 5:72-79

  6. ^ von Herrath M. et al. Type 1 diabetes as a relapsing-remitting disease? Nat Rev Immunol 2007; 7:988-994

  7. ^ Sosenko J.M. et al. Glucose and C-peptide changes in the peri-onset period of type 1 diabetes in the diabetes prevention trial-type 1. Diabetes Care 2008; 31:2188-2192

  8. ^ Srikanta S. et al. First-degree relatives of patients with type I diabetes mellitus. Islet-cell antibodies and abnormal insulin secretion. N Engl J Med 1985; 313:461-464

  9. ^ Vardi P. et al. Predictive value of intravenous glucose tolerance test insulin secretion less than or greater than the first percentile in islet cell antibody positive relatives of type 1 (insulin-dependent) diabetic patients. Diabetologia 1991; 34:93-102

  10. ^ Bingley P.J. Interactions of age, islet cell antibodies, insulin autoantibodies, and first-phase insulin response in predicting risk of progression to IDDM in ICA+ relatives: the ICARUS data set. Islet Cell Antibody Register Users Study. Diabetes 1996; 45:1720-1728

  11. ^ Vandemeulebroucke E. et al. Hyperglycaemic clamp test for diabetes risk assessment in IA-2-antibody-positive relatives of type 1 diabetic patients. Diabetologia 2010; 53:36-44


Nobody has commented on this article

Commenting is only available for registered Diapedia users. Please log in or register first.