Mixed‐meal tolerance test to assess residual beta‐cell secretion: Beyond the area‐under‐curve of plasma C‐peptide concentration

Aims Residual beta‐cell secretion in type 1 diabetes is commonly assessed by area‐under‐curve of plasma C‐peptide concentration (AUCCpep) following mixed‐meal tolerance test (MMTT). We aimed to investigate alternative measures of beta‐cell responsiveness. Methods We analyzed data from 32 youth (age 7 to 17 years) undergoing MMTT within 6 months of type 1 diabetes diagnosis. We related AUCCpep with (a) validated mechanistic index of postprandial beta‐cell responsiveness M I accounting for glucose level during MMTT, and (b) pragmatic marker calculated as baseline plasma C‐peptide concentration corrected for baseline plasma glucose concentration. Results Postprandial responsiveness M I was correlated with age and BMI SDS (R s = 0.66 and 0.44, P < 0.01 and P < 0.05) and was more correlated with glycated hemoglobin than AUCCpep (R s = 0.79, P = 0.04). The pragmatic marker was highly correlated with AUCCpep (R s = 0.94, P < 0.01). Conclusions Postprandial responsiveness M I may be more relevant to glucose control than AUCCpep. Baseline C‐peptide corrected for baseline glucose appears to be a suitable surrogate of AUCCpep if MMTT is not performed.


| INTRODUCTION
The area-under-curve of sequential C-peptide concentrations (AUC Cpep ) during the mixed-meal tolerance test (MMTT) is the goldstandard method to assess residual beta-cell (ie, insulin) secretion in type 1 diabetes. 1 Traditionally, glucose excursions during the MMTT are not taken into account, although these impact on the magnitude of C-peptide response. 2 In this work, we re-analyzed MMTT data obtained in newly diagnosed children and adolescents with type 1 diabetes aged 7 to of AUC Cpep and (b) to explore the relationships among demographic and clinical factors, and AUC Cpep and its surrogate markers.

| METHODS
We analyzed data obtained from 32 participants with newly diagnosed type 1 diabetes (age 12.4 [2.9] years, 12 males, HbA1c 6.8% All participants (aged ≥16 years) gave informed consent, and children <16 years gave assent and their parents gave informed consent to the study procedures.
The MMTT was performed following an overnight fast, with no food or drink other than water from midnight, and at baseline glucose levels between 4 and 11.1 mmol/L. Long-acting insulin and basal rates for insulin pump users were continued as normal. The use of rapidacting insulin bolus was acceptable up to 2 hours before the MMTT and the use of short-acting insulin bolus up to 6 hours before the MMTT. Participants ingested 6 mL/kg of Boost meal solution (maximum 360 mL), within 10 minutes. Blood samples for the measurement of C-peptide and glucose were collected 10 minutes prior to the meal (−10 minutes), at the time of ingestion (0 minutes), and at 15, 30, 60, 90 and 120 minutes.  Figure S1 shows a sample model fit to measured plasma Cpeptide including measurements of plasma glucose (the forcing function). Figure S2 depicts weighted residuals across all participants demonstrating acceptable fit of the model to plasma C-peptide measurements. Table 1

| DISCUSSION
The present analysis demonstrates the feasibility of using a model of C-peptide kinetics to assess residual beta-cell function during MMTT in newly-diagnosed type 1 diabetes. Traditionally, the AUC Cpep during a MMTT has not been corrected for glucose excursions, which are likely to affect the amplitude of the C-peptide response. 2 8 Data from the present analysis suggest that baseline C-peptide corrected for baseline glucose may be a more appropriate marker than baseline C-peptide and a more cost-effective marker than the stimulated C-peptide concentration sidestepping the need for MMTT and complexity of the assessment.
We show that M I was more tightly correlated with HbA1c than AUC Cpep (P = 0.04) indicating that M I may be a more clinically relevant marker of C-peptide secretion than AUC Cpep . The study is limited by a relatively small sample size. Further analyses with larger datasets and longitudinal evaluations are warranted. We applied parameters of Cpeptide kinetics determined in healthy subjects. As C-peptide is eliminated primarily by the kidney and assuming comparable kidney function among healthy individuals and those with recently diagnosed type 1 diabetes, we consider this limitation to be of little significance to our findings.
Alternative C-peptide secretion models assume a more complex relationship between glucose concentration and insulin secretion compared to the model used in the present study. 9,10 These alternative models may provide additional information about C-peptide secretory characteristics but require more frequent sampling. Our  parameter M I represents the dominant relationship between plasma glucose and C-peptide secretion and is accounted for in the alternative approaches. 9,10 In conclusion, baseline C-peptide corrected for baseline glucose may be a suitable surrogate marker of residual beta-cell in newlydiagnosed type 1 diabetes. Postprandial pancreatic responsiveness estimated through a model of C-peptide kinetics appears more relevant to glucose control than the conventional area-under-curve of plasma C-peptide concentration following MMTT.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of the article.