I~-Cell Function During Insulin-Modified Intravenous Glucose Tolerance Test
Successfully Assessed by the C-Peptide Minimal Model
Gianna Toffolo, William T. Cefalu, and Claudio Cobelli
The insulin-modified intravenous glucose tolerance test (IM-IVGTT} is increasingly used to measure insulin sensitivity.
However, the assessment of I~-cell secretion is usually made using rough indices. The aim here is to evaluate the ability of the
minimal model of C-peptide secretion and kinetics recently proposed for the standard IVGTT (S-IVGTT) to also assess I~-cell
function during the IM-IVGTT. C-peptide and glucose data from the IM-IVGTT in 15 normal humans were analyzed. The results
show that the same rich i~-cell picture from the S-IVGTT can be obtained during an IM-IVGTT. In particular, in each individual,
the time course of I~-cell secretion can be reconstructed and the functional indices of glucose control on first-phase (O1),
second-phase (O2), and basal (~b) insulin secretion can be estimated (O1 = 191 _+ 29, O2 = 10.9 +__ 1.4.10 -9 • min -1, and
(I) b = 5.7 _+ 1.0.10 -9 • min -1, mean -+ SE). Finally, the comparison between IM-IVGTr and S-IVGTT O1, Oz, and Ob values
suggest they are not affected by insulin administration.
Copyright © 1999 by W.B. Saunders Company
T
HE INSULIN-MODIFIED intravenous glucose tolerance
test (IM-IVGTT) 1.2 is increasingly used 3 to obtain a
quantitative picture of glucose disposal through the minimal
model indices of insulin sensitivity and glucose effectiveness. 4
In contrast, the current IM-IVGTT assessment of ~-cetl func-
tion is rather approximate, being usually based on the acute
insulin response to glucose (AIR) index, ie, the area under the
initial portion of insulin data before insulin administration,
which reflects not only first-phase secretion but also hepatic
insulin extraction and the whole-body insulin clearance rate.
This approximate IM-IVGTT description appears somewhat at
odds with the description obtained by the standard IVGTT
(S-1VGTT) in which C-peptide is measured, and several
methods can be applied to obtain a rich parametric portrait of
13-cell function) -s In particular, the C-peptide minimal model
originally proposed by our group 8 and applied in several
studies TM allows a reconstruction of the prehepatic insulin
secretion profile and quantification of the indices of glucose
control on first-phase, second-phase, and basal insulin secre-
tion,
The aim herein is to investigate if a detailed portrait of fJ-cell
function can also be obtained from an IM-IVGTT if glucose and
C-peptide concentrations are interpreted with the S-IVGTT
minimal model. 8
SUBJECTS AND METHODS
Subjects
Studies were performed in 15 healthy, community-dwelling subjects
(age, 51 -+ 4 years; weight, 84 ± 5 kg; body mass index, 30.3 +_ 1.6
kg~m2). Participants underwent a standardized interview to obtain
demographic data and information documenting eligibility. All subjects
were fully ambulatory and normally active and used no medications
From the Department of Electronics and Informatics, University of
Padova, Padova, Italy; and Department of Medicine, University of
Vermont, Burlington, WE.
Submitted November 10, 1998; accepted March 29, 1999.
Supported by a grant from the Italian Ministero dell'Umversitgt e
della Rieerca Scientifica e Tecnologica (MURST 40%) on "Biosistemi e
Bioinformatiea. '"
Address reprint requests to Claudio Cobelli. PhD, Dipartimento di
Elettronica e Informatiea, Via Gradenigo 6a, 35131 Padova, Italy.
Copyright © 1999 by W.B. Saunders Company
0026-0495/99/4809-0015510.00/0
known to affect glucose metabolism, blood pressure, or lipids. Partici-
pants were considered healthy as determined by a medical history,
physical examination, electrocardiogram, complete blood cell count,
routine blood and urine chemistry analysis, and thyroid function tests.
All subjects provided written informed consent to participate in the
study, which was approved by the Clinical Research Practices Commit-
tee of the Bowman Gray School of Medicine.
Protocol
The IM-IVGTT study was initiated in the morning after an overnight
fast. Two 18-gauge intravenous catheters were placed in each forearm
and kept patent by controlled-flow saline infusion. Each line was
equipped with a three-way stopcock. One line was used for intravenous
administration of glucose and insulin, and the other was used to obtain
blood samples. Blood samples of 1 mL for insulin and glucose were
drawn from the line at 15 minutes, 5 minutes, and immediately before
glucose (300 mg/kg) was injected intravenously, and the line was
flushed with saline solution. Insulin (0.03 U/kg) was infused 20 to 25
rmnutes after glucose injection. Blood samples of 1 mL were drawn at 2,
3, 4, 5, 8, 10, 12, 14, 16, 18, 20, 22, 24, 28, 32, 40, 45, 50, 60, 70, 80, 90,
100, 110, 120, 140, 160, and 180 minutes after glucose injection. The
samples were centrifuged immediately, and the plasma was placed on
ice. Glucose deterrmnations were performed immediately after centrifu-
gation using the glucose oxidase method on a Glucose Analyzer 2
(Beckman, Brea, CA: mtraassay coefficient of variation [CV]. 2%).
C-peptade was assayed from frozen plasma by radioimmunoassay
(IncStar, Stillwater, MN; intraassay CV, 6%).
C-Peptide Kinetic Model
IM-IVGTT C-peptxde and glucose data were analyzed with the
C-peptide minimal model. 8 We will summarize its salient features and
describe the parameters it provides. The model is shown in Fig 1 and is
described by the following:
CPI(t) = - [k01 + k2dCPl(t) + k12CP2(t)+ mX(O
CP2(t ) = k21CPl(t ) -- k12CP2(t )
X(t) = - rnX(t) + Y(t)
Y(t) = - e~ [Y(t) - 13[a(t) - h]}
CPI(O) = 0
CP2(O) = 0
Eq 1
x(o) = xo
Y(O) = o.
CP1 and CP2 (picomolars) are the C-peptide concentration (above basal)
in the accessible and peripheral compartments, respectively: X (picomo-
lars) and Y (picomoles per liter per minute) are, respectively, the
C-peptide amount and provision in the 13cells, both normalized to the
distribution volume of compartment 1; k u (per minute) values are
1162 Metabolism, Vol 48, No 9 (September), 1999: pp 1162-1166