Lessons from the mixed-meal tolerance test: use of 90-minute and fasting C-peptide in pediatric diabetes

Abstract

Objective: Mixed-meal tolerance test (MMTT) area under the curve C-peptide (AUC CP) is the gold-standard measure of endogenous insulin secretion in type 1 diabetes but is intensive and invasive to perform. The 90-min MMTT-stimulated CP ≥0.2 nmol/L (90CP) is related to improved clinical outcomes, and CP ≥0.1 nmol/L is the equivalent fasting measure (FCP). We assessed whether 90CP or FCP are alternatives to a full MMTT.

Research design and methods: CP was measured during 1,334 MMTTs in 421 type 1 diabetes patients aged <18 years at 3, 9, 18, 48, and 72 months duration. We assessed: 1) correlation between mean AUC CP and 90CP or FCP; 2) sensitivity and specificity of 90CP ≥0.2 nmol/L and FCP ≥ 0.1 nmol/L to detect peak CP ≥0.2 nmol/L and the equivalent AUC CP; and 3) how the time taken to reach the CP peak varied with age of diagnosis and diabetes duration.

Results: AUC CP was highly correlated to 90CP (r(s) = 0.96; P < 0.0001) and strongly correlated to FCP (r(s) = 0.84; P < 0.0001). AUC CP ≥23 nmol/L/150 min was the equivalent cutoff for peak CP ≥0.2 nmol/L (98% sensitivity/97% specificity). A 90CP ≥0.2 nmol/L correctly classified 96% patients using AUC or peak CP, whereas FCP ≥0.1 nmol/L classified 83 and 85% patients, respectively. There was only a small difference seen between peak and 90CP (median 0.02 nmol/L). The CP peak occurred earlier in patients with longer diabetes duration (6.1 min each 1-year increase in duration) and younger age (2.5 min each 1-year increase).

Conclusions: 90CP is a highly sensitive and specific measure of AUC and peak CP in children and adolescents with type 1 diabetes and offers a practical alternative to a full MMTT.

Figures

Figure 1

Correlation between mean AUC CP and mean 90-min–stimulated serum CP (rs = 0.96) (A) and mean FCP (rs = 0.84) (B) during a 150-min MMTT (n = 421).

Figure 2

Timing and magnitude of the CP peak during 150-min MMTT, split by age of diagnosis tertiles (n = 141; 10–13 years [light gray bar], n = 139; and ≥13 years [dark gray bar], n = 141). A: The CP peak occurs earlier with increasing diabetes duration and younger age of diagnosis (1,334 MMTTs in 421 patients). B: Peak CP values reduce with increasing diabetes duration and are lower in younger patients (1,334 MMTTs in 421 patients).

Figure 3

Kaplan-Meier survival plots to show the impact of FCP (A) and age at diagnosis (B) on the time taken for patients to become insulin deficient (peak CP <0.2 nmol/L). A: FCP is divided by tertiles, in which dashed black line refers to FCP <0.17 nmol/L, solid black line refers to FCP 0.17–0.29 nmol/L, and dashed gray line refers to FCP ≥0.29 nmol/L. Current duration is used in the cases of censored data (+) in which last recorded peak CP is >0.2 nmol/L). B: Age of diagnosis is divided by tertiles, in which dashed black line refers to age <10 years, solid black line refers to age 10–13 years, and dashed gray line refers to age ≥13 years. Current duration is used in the cases of censored data (+) in which last recorded peak CP is >0.2 nmol/L).