Received: 13 June 2019 Revised: 4 October 2019 Accepted: 26 October 2019

DOI: 10.1111/pedi.12940

ORIGINAL ARTICLE: OBESITY/INSULIN RESISTANCE,

TYPE 2 DIABETES

β-cell function, incretin response, and insulin sensitivity

of glucose and fat metabolism in obese youth: Relationship
to OGTT-time-to-glucose-peak

Joon Young Kim1 | Hala Tfayli2 | Fida Bacha3 | SoJung Lee4 |

Sara F. Michaliszyn5 | Shahwar Yousuf6 | Nour Gebara6 | Silva Arslanian6,7

1
Department of Exercise Science, Syracuse
University, Syracuse, New York Abstract
2
Department of Pediatrics and Adolescent Background: In adults, the time-to-glucose-peak at or after 30 minutes during an oral
Medicine, American University of Beirut
glucose tolerance test (OGTT) identifies physiologically distinct groups with differ-
Medical Center, Beirut, Lebanon
3
Children's Nutrition Research Center, Baylor ences in insulin sensitivity, β-cell function and risk for type 2 diabetes. In obese non-
College of Medicine, Houston, Texas diabetic adolescents, we investigated if the OGTT-time-to-glucose-peak also reflects
4
Division of Sports Medicine, Graduate School
incretin and free fatty acid (FFA) responses besides insulin sensitivity and β-cell func-
of Physical Education, Kyung Hee University,
Yongin, Republic of Korea tion, measured by the clamp.
5
Kinesiology and Sport Science, Youngstown Methods: Obese adolescents (n = 278) were categorized according to their OGTT-
State University, Youngstown, Ohio
6
time-to-glucose-peak by Early-peak (at 30 minutes) vs Late-peak (>30 minutes)
Center for Pediatric Research in Obesity and
Metabolism, UPMC Children's Hospital of groups. Body composition, visceral adipose tissue, oral disposition index and
Pittsburgh, Pittsburgh, Pennsylvania OGTT-area under the curve (AUC) were examined. A subset of 102 participants
7
Division of Pediatric Endocrinology,
had both hyperinsulinemic-euglycemic and hyperglycemic clamps to measure
Metabolism and Diabetes Mellitus, UPMC
Children's Hospital of Pittsburgh, Pittsburgh, in vivo insulin sensitivity, insulin secretion, and β-cell function relative to insulin
Pennsylvania
sensitivity.
Correspondence Results: Compared with the Early-peak group, the Late-peak group had impaired
Silva Arslanian, MD, Division of Pediatric
β-cell function relative to insulin sensitivity, lower glucose-dependent insulinotropic
Endocrinology, Metabolism and Diabetes
Mellitus, UPMC Children's Hospital of polypeptide-AUC, and higher FFA-AUC despite higher insulin- and C-peptide-AUC.
Pittsburgh, 4401 Penn Avenue, Pittsburgh,
They also had lower hepatic and peripheral insulin sensitivity despite similar percent
PA 15224.
Email: silva.arslanian@chp.edu body fat and visceral adipose tissue, and had higher prevalence of impaired glucose
tolerance (all P < .05).
Funding information
American Diabetes Association, Grant/Award Conclusions: In obese non-diabetic youth, those with a Late-peak vs an Early-peak
Number: 7-08-JF-27; National Center for
glucose during an OGTT showed diminished β-cell function, blunted incretin secre-
Advancing Translational Sciences, Grant/
Award Number: UL1TR000005; National tion, and lower insulin sensitivity of glucose and FFA metabolism. It remains to be
Center for Research Resources, Grant/Award
determined if Late-peak glucose predicts the future development of type 2 diabetes
Number: UL1RR024153; National Institute of
Child Health and Human Development, Grant/ in these high-risk youth.
Award Numbers: K24-HD01357,
R01-HD27503; National Institute of Diabetes
KEYWORDS
and Digestive and Kidney Diseases, Grant/
Award Number: R21DK083654-01A1 clamp, glucose peak, insulin sensitivity, obesity, OGTT, youth, β-cell function

© 2019 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd

18 wileyonlinelibrary.com/journal/pedi Pediatric Diabetes. 2020;21:18–27.

KIM ET AL.
1 | I N T RO DU CT I O N
The oral glucose tolerance test (OGTT) is a screening tool for glucose
intolerance identifying diabetes and prediabetes, which includes
impaired fasting glucose (IFG) and impaired glucose tolerance (IGT), a
state of heightened risk for future type 2 diabetes.1 However, use of
single cut-points for IFG and IGT to predict the development of type
2 diabetes has been limited as prospective epidemiological studies in
adults demonstrated that only one-half of prediabetic individuals
eventually convert to type 2 diabetes.2,3 For this reason, there is
increasing interest in finding novel biomarkers derived from the OGTT
that can identify metabolic abnormalities beyond fasting and 2-hour
glucose concentrations. Such emerging risk indicators include OGTT-
glucose-response-curve,4-8 1-hour glucose concentration,9-11 and
OGTT-time-to-glucose-peak (at 30 minutes vs >30 minutes).8,12-14
A recent study in normoglycemic adults tested reproducibility of
the aforementioned markers and suggested that the timing of the glu-
cose peak during an OGTT shows higher intra-subject reproducibility
(k coefficient 0.76) compared with other indicators including glucose-
response-curve (k coefficient 0.35) or 1-hour glucose concentration
(k coefficient 0.64).15 As a result, the OGTT-time-to-glucose-peak has
been considered a reliable indicator of type 2 diabetes risk.8,12-17 In
adults, delayed timing of the glucose peak (occurring after 30 minutes
during an OGTT) compared with Early-peak (occurring at 30 minutes)
is associated with lower insulin sensitivity, impaired insulin secretion,
and higher risk for prediabetes and type 2 diabetes.8,12,13,16 However,
it remains unknown if the OGTT-time-to-glucose-peak reflects dis-
tinct responses in incretin hormones and free fatty acid (FFA) concen-
trations; especially since incretins and adipose insulin resistance are
important pathophysiological components of type 2 diabetes.18-21
Additionally, most studies have not used sensitive measures of in vivo
insulin sensitivity or β-cell function, rather used OGTT-derived surro-
gate estimates.12,15-17 Furthermore, data are very sparse in
pediatrics.14
Therefore, the purpose of this study was to investigate differ-
ences in (a) clamp-measured biomarkers of type 2 diabetes (ie, hepatic
and peripheral insulin sensitivity, and β-cell function), (b) adipose tis-
sue insulin resistance, and (c) OGTT-incretin responses by the OGTT-
time-to-glucose-peak, Early-peak (at 30 minutes) vs Late-peak
(>30 minutes).
2 | METHODS
2.1 | Participants
Data from 278 non-diabetic adolescents (138 African American,
140 American White; ages 10 to <20 years; Tanner stages II-V), who
participated in our National Institutes of Health-funded K24 Grant of
Childhood Insulin Resistance, from January 2004 to October 2012,
were examined. Data resulting from this grant but unrelated to the
OGTT-time-to-glucose-peak have been published. 7,22-25
There were
29 who were overweight with BMI ≥85th percentile for age and sex
19
but <95th and 249 obese with BMI ≥95th percentile. From here on
we will refer to the overweight and obese youth as “obese” for sim-
plicity purposes. There were 214 participants with normal glucose tol-
erance (NGT), and 64 with IGT. Participants were recruited through
newspaper advertisements, flyers posted in the medical campus, city
bus routes and the outpatient clinics in the Weight Management and
Wellness Center and the Division of Pediatric Endocrinology. The
study was approved by the Institutional Review Board of the Univer-
sity of Pittsburgh, and written informed parental consent and child
assent were obtained from all participants before any research proce-
dures in accordance with the ethical guidelines of Children's Hospital
of Pittsburgh.
2.2 | Procedures
All procedures were performed at the Pediatric Clinical and Transla-
tional Research Center (PCTRC) of Children's Hospital of Pittsburgh.
All participants underwent medical history, physical examination, and
hematologic and biochemical tests. Height and weight were assessed
to the nearest 0.1 cm and 0.1 kg, respectively, and used to calculate
body mass index (BMI). Pubertal development was assessed using
Tanner criteria.26 Body composition was evaluated with DEXA for the
measurement of total body fat mass and percent body fat. Abdominal
visceral adipose tissue was assessed by either magnetic resonance
imaging (MRI) or computed tomography (CT) at L4-5 intervertebral
space.7,27 The switch from CT to MRI was imposed by the study
section during the competitive grant renewal process. Note that
Klopfenstein et al28 demonstrated strong correlation (r = 0.89-0.95)
and good agreement (ie, very small average difference) between CT
and MRI for the measurement of abdominal adipose tissue.
2.3 | Metabolic studies
After 10 to 12 hours of overnight fasting, participants underwent a
2-hour OGTT (1.75 g/kg, maximum 75 g).7,24 Blood samples were
obtained at −15, 0, 15, 30, 60, 90 and 120 minutes for the measure-
ment of glucose, insulin, C-peptide, FFA, glucose-dependent
insulinotropic polypeptide (GIP), total glucagon-like peptide 1 (GLP-1),
and glucagon. Fasting blood samples were obtained for the determi-
nation of lipid profile and HbA1c.
Of the total 278 participants, either the day after the OGTT or on
a separate visit within a 1- to 4-week period, a subset of 102 partici-
pants (68 NGT and 34 IGT) were admitted twice to the PCTRC for a
hyperinsulinemic-euglycemic clamp to assess in vivo insulin sensitiv-
ity, and a hyperglycemic clamp to assess insulin secretion, performed
in random order.29 Each clamp evaluation was performed after a 10-
to 12-hour overnight fast. Fasting hepatic glucose production was
measured before the start of the hyperinsulinemic-euglycemic clamp,
with a primed (2.2 μmol/kg) constant infusion of [6,6-2H2]glucose at
0.22 μmol/kg/min for a total of 2 hours as described.29 After the
2-hour baseline isotope infusion period, in vivo insulin sensitivity was
20
evaluated during a 3-hour hyperinsulinemic (80 mu/m2/min)-
euglycemic clamp (100 mg/dL).29 First- and second-phase insulin
secretions were assessed during a 2-hour hyperglycemic (225 mg/dL)
clamp as described before.29 Plasma glucose was increased rapidly to
225 mg/dL by a bolus infusion of dextrose and maintained at that
level by a variable rate infusion of 20% dextrose for 2 hours, with fre-
quent measurement of glucose and insulin concentrations.
2.4 | Biochemical measurements
During the 2-hour OGTT, blood was collected in chilled aprotinin/
EDTA tubes for insulin, C-peptide and glucagon measurements.
Dipeptidyl peptidase-4 inhibitor (10 μL, Catalog no. DPP4; Millipore,
St. Charles, Missouri) was added before sampling to the aprotinin/
EDTA tubes to prevent the enzymatic degradation of FFA, GIP, and
total GLP-1. Blood samples were immediately separated in a refriger-
ated centrifuge. Plasma samples were divided into aliquots and stored
at −80 C until analysis. Plasma glucose was determined by the glu-
cose oxidase method using a glucose analyzer (Yellow Springs Instru-
ment Co., Yellow Springs, Ohio), and insulin, C-peptide and glucagon
by commercially available radioimmunoassay (Millipore), as previously
7
reported. FFA was determined using enzymatic colorimetric methods
with a Wako (non-esterified fatty acids) NEFA-HR test kit (Wako,
Osaka, Japan). GIP was measured on the Luminex 200 IS (Luminex,
Austin, Texas) using a two-plex human gut hormone MILLIPLEX kit
(Catalog no. HGT-68K-02; Millipore) as described before.21 Total
GLP-1 was measured on a microplate reader (BioTek, Winooski, Ver-
mont) using a multispecies total GLP-1 ELISA kit (Catalog
no. EZGLP1T-36K; Millipore).21 HbA1c was measured by high-
performance liquid chromatography (Tosoh Medics, Inc., San Fran-
cisco, California). Plasma lipid concentrations were determined using
the standards of the Centers for Disease Control and Prevention.29
2.5 | Classification of Early-peak vs Late-peak
The time-to-glucose-peak was defined as the time point when the glu-
cose concentration was highest (30, 60, 90 or 120 minutes) during an
OGTT. Youth with a glucose peak occurring at 30 minutes or earlier
are classified as an “Early-peak” group while those with a glucose peak
occurring after 30 minutes (ie, 60, 90 or 120 minutes) were catego-
rized as a “Late-peak” group.
2.6 | Calculations
OGTT-area under the curve (AUC) for glucose, insulin, C-peptide,
FFA, GIP, total GLP-1, and glucagon was calculated with the use of
the trapezoidal method. For the total cohort, insulin sensitivity was
estimated by the homeostasis model assessment for insulin sensitivity
(HOMA-IS) which is the reciprocal of HOMA-IR (HOMA2 Calculator
at http://www.dtu.ox.ac.uk). Adipose tissue insulin resistance
KIM ET AL.
(Adipose-IR) was calculated as fasting insulin × fasting FFA concentra-
tions.20,30 Early-phase insulin secretion during the OGTT (ie,
insulinogenic index) was calculated as: Δinsulin (0-30 minutes)/
Δglucose (0-30 minutes) as previously described.24 Oral disposition
index (oDI) which represents β-cell function relative to insulin sensitiv-
ity was calculated as the product of HOMA-IS × insulinogenic
index.24
For the subset of participants who underwent both clamp proce-
dures, fasting hepatic glucose production was calculated during the
last 30 minutes of the baseline 2-hour isotope infusion (−30 to
0 minutes) according to steady-state tracer dilution equations.29
Hepatic insulin sensitivity was calculated as the inverse of the product
of hepatic glucose production and fasting plasma insulin concentra-
tion.29,31 Insulin-stimulated glucose disposal (Rd) was calculated to be
equal to the rate of exogenous glucose infusion during the final
30 minutes of the hyperinsulinemic-euglycemic clamp. Peripheral
insulin sensitivity was calculated by dividing the Rd by the steady-
state clamp insulin concentration multiplied by 100.29 During the
hyperglycemic clamp, first- and second-phase insulin concentrations
were calculated as before.25,29 β-cell function relative to insulin sensi-
tivity, that is, the disposition index (DI), was calculated as the product
of insulin sensitivity, measured by the hyperinsulinemic-euglycemic
clamp, and first-phase insulin secretion measured during the hypergly-
cemic clamp.22,23,25,29
2.7 | Statistical analysis
Independent sample t tests and chi-square analyses were used to
compare physical and metabolic characteristic between the Early-peak
(glucose peak at 30 minutes) and the Late-peak (glucose peak
>30 minutes) groups. Analysis of covariance was used to compare
groups after adjusting for potential confounding effects (BMI, total
body fat mass, and glycemic status). Repeated-measures analysis of
variance was used to assess differences in OGTT parameters at each
time point during the test. Data that did not meet the assumptions for
normality were log10 transformed; untransformed data are presented
for ease of interpretation. Data were analyzed using PASW 21.0 sta-
tistical software package with significance set at P ≤ .05. Data are
mean ± SEM.
3 | RE SU LT S
3.1 | Physical and metabolic characteristics of the
total OGTT cohort: comparison of Early-peak vs Late-
peak groups
Based on the time-to-glucose-peak during the 2-hour OGTT, 142 par-
ticipants (51%) were classified as Early-peak (by 30 minutes); 4 at
15 minutes and 138 at 30 minutes, and 136 (49%) were categorized
as Late-peak (>30 minutes); 107 at 60 minutes, 19 at 90 minutes, and
10 at 120 minutes (Figure S1). Physical and metabolic characteristics
KIM ET AL. 21

TABLE 1 Physical and metabolic characteristics of 278 participants grouped into Early-peak vs Late-peak based on OGTT-glucose-time-
to-peak

Variables Early-peak (n = 142) Late-peak (n = 136) P-value

Physical characteristics
Age (y) 14.7 ± 0.2 14.9 ± 0.2 .378
Race (n, AA/AW) 68/74 70/66 .631
Sex (n, male/female) 59/83 43/93 .106
Tanner (n, II/III/IV/V) 5/19/19/99 2/13/17/104 .472
BMI (kg/m2) 34.1 ± 0.6 36.3 ± 0.6 .006
BMI percentile 97.5 ± 0.2 97.9 ± 0.2 .247
Total body fat mass (kg) 39.7 ± 1.1 43.8 ± 1.2 .022
Percent body fat (%) 42.3 ± 0.6 43.8 ± 0.6 .157
Visceral adipose tissue (cm2) 65.4 ± 2.8 71.5 ± 3.5 .190
Metabolic characteristics
Glycemic status (n, NGT/IGT) 125/17 89/47 <.0001
HbA1c (%) 5.44 ± 0.06 5.44 ± 0.04 .935
HbA1c (mmol/mol) 35.91 ± 0.69 35.92 ± 0.45
Cholesterol (mg/dL) 152.5 ± 2.9 149.8 ± 2.5 .654
Triglyceride (mg/dL) 96.8 ± 4.6 101.5 ± 4.9 .205
HDL (mg/dL) 50.2 ± 2.3 47.0 ± 1.8 .272
LDL (mg/dL) 86.1 ± 2.7 83.1 ± 2.6 .496
VLDL (mg/dL) 19.7 ± 0.9 19.6 ± 0.9 .619
Fasting- and OGTT-based parameters
Fasting glucose (mg/dL) 89.7 ± 0.6 89.4 ± 0.7 .578
Fasting insulin (μU/mL) 25.9 ± 1.1 34.3 ± 2.0 .004a
Fasting C-peptide (ng/mL) 2.8 ± 0.1 3.4 ± 0.1 .006a
2-hour glucose (mg/dL) 118.5 ± 1.4 130.2 ± 2.0 <.0001a,b
2-hour insulin (μU/mL) 140.0 ± 9.9 217.8 ± 14.3 <.0001a,b
2-hour C-peptide (ng/mL) 9.1 ± 0.3 12.0 ± 0.6 <.0001a,b
Glucose AUC (mg/dL/h) 14 741.9 ± 145.3 15 959.0 ± 193.3 <.0001a,b
Insulin AUC (μU/mL/h) 18 726.6 ± 1023.4 22 251.9 ± 1210.1 .048
C-peptide AUC (ng/mL/h) 1048.5 ± 36.4 1211.1 ± 42.6 .005a
FFA AUC (mmol/L/h) 16.9 ± 0.7 19.9 ± 0.8 .003a,b
GIP AUC (pg/mL/h) 22 625.9 ± 770.2 19 700.9 ± 668.3 .006a,b
Total GLP-1 AUC (pmol/L/h) 1702.0 ± 100.4 1770.8 ± 101.7 .465
Glucagon AUC (pg/mL/h) 6473.6 ± 210.0 6493.2 ± 217.9 .816
HOMA-IS 0.23 ± 0.01 0.20 ± 0.01 .006a
Insulinogenic index (μU/mL/mg/dL) 4.38 ± 0.28 3.69 ± 0.27 .043a,b
oDI (HOMA-IS × insulinogenic index) 0.86 ± 0.06 0.61 ± 0.05 <.0001a,b
Adipose-IR (μU/mL × mmol/L) 7.2 ± 0.6 10.0 ± 0.7 <.0001a,b

Note: Values are mean ± SEM. Visceral adipose tissue data were available on 233 (126 vs 107). Total n for FFA is 251 (140 vs 121), for GIP is 252 (129 vs
123), for total GLP-1 is 244 (128 vs 116), for glucagon is 275 (141 vs 134).
Abbreviations: AA, African American; AUC, area under the curve; AW, American White; BMI, body mass index; FFA, free fatty acid; GIP, glucose-
dependent insulinotropic polypeptide; GLP-1, total glucagon-like peptide 1; HDL, high-density lipoprotein; HOMA-IS, homeostasis model assessment for
insulin sensitivity; IGT, impaired glucose tolerance; LDL, low-density lipoprotein; NGT, normal glucose tolerance; NS, not significant; oDI, oral disposition
index; OGTT, oral glucose tolerance test; VLDL, very low-density lipoprotein.
a
P < .05 after adjusting for BMI and total body fat mass.
b
P < .05 after adjusting for glycemic status, BMI, and total body fat mass.
22
of all participants categorized into the Early-peak vs the Late-peak
groups are presented in Table 1. The two groups were similar in
age, race, sex, Tanner stage, BMI percentile, percent body fat, vis-
ceral adipose tissue and lipid profile, and different in BMI and total
body fat mass, higher in the Late-peak vs the Early-peak group
(Table 1).
Despite similar HbA1c and fasting glucose, the prevalence of IGT
was higher in the Late-peak (35%) vs the Early-peak (12%) group
(Table 1). Participants in the Late-peak group exhibited higher 2-hour
glucose, higher fasting and 2-hour insulin and C-peptide concentra-
tions, lower HOMA-IS, lower insulinogenic index, lower oDI, and
higher Adipose-IR compared with those in the Early-peak group
(Table 1). With respect to OGTT-AUC, the Late-peak group had higher
glucose-AUC and FFA-AUC, despite higher insulin- and C-peptide-
AUCs. Further, the Late-peak group had lower GIP-AUC compared
with the Early-peak group with no differences in total GLP-1 and glu-
cagon AUCs (Table 1). All significant differences in metabolic parame-
ters, with the exception of insulin-AUC, were independent of BMI and
total body fat mass (Table 1). Similar results were obtained after
adjusting for glycemic status in addition to BMI and total body fat
mass (Table 1). Furthermore, when the dataset was restricted to only
those participants with NGT, the Late-peak group had significantly
higher glucose-AUC, 2-hour insulin and C-peptide, fasting FFA and
FFA-AUC, Adipose-IR and lower fasting and 2-hour GIP and GIP-
AUC, and oDI compared with the Early-peak group (data not shown).
F I G U R E 1 Glucose (A), insulin (B), C-peptide (C), FFA (D), GIP (E) and total GLP-1 (F) responses during a 2-hour OGTT in Early-peak (open
squares and dashed lines) and Late-peak (filled circles and solid lines) groups. For simplicity purposes, the 15 minutes time point of the OGTT is
not shown. *P < .05, **P < .01, ***P < .001
KIM ET AL.
Figure 1A-F displays the OGTT response for glucose (1A), insulin (1B),
C-peptide (1C), FFA (1D), GIP (1E), and total GLP-1 (1F) in each of the
Early-peak vs the Late-peak groups.
3.2 | Clamp measured in vivo insulin sensitivity,
insulin secretion and β-cell function relative to insulin
sensitivity in Early-peak vs Late-peak groups
Of the 102 participants with all three tests, the OGTT, the
hyperinsulinemic-euglycemic and the hyperglycemic clamps, 49 (48%)
were classified into the Early-peak group and 53 (52%) into the Late-
peak group. The pattern of differences in physical and metabolic char-
acteristics between the two groups was similar to the total cohort:
elevated prevalence of IGT, and higher 2-hour glucose, glucose-AUC
and 2-hour insulin in the Late-peak compared with the Early-peak
groups before and after adjustment for BMI and total body fat mass
(Table S1). After adjusting for glycemic status in addition to BMI and
total body fat mass, fasting GIP concentration was still significantly
lower in the Late-peak vs the Early-peak (Table S1).
When the Early-peak vs the Late-peak groups were compared
with respect to clamp-based parameters, participants in the Late-peak
group had lower hepatic and peripheral insulin sensitivity (Figure 2A,
B). β-cell function relative to insulin sensitivity, that is, DI was signifi-
cantly lower in the Late-peak vs the Early-peak groups (Figure 3A),
KIM ET AL. 23

F I G U R E 2 In vivo hepatic insulin sensitivity (A) and peripheral insulin sensitivity (B) in Early-peak vs Late-peak groups. Adjusted P is for the
difference after adjusting for BMI and total body fat mass

F I G U R E 3 β-cell function relative to insulin sensitivity (disposition index) (A), and the hyperbolic relationship between insulin sensitivity
(measured by the hyperinsulinemic-euglycemic clamp) and first-phase insulin (B) (measured by the hyperglycemic clamp) in Early-peak (open
squares and dashed lines) vs Late-peak (filled circles and solid lines). The hyperbolic curve is the nonlinear fitted regression line derived from the
individual data points of the Early-peak (R2 = 0.360, P < .0001) vs the Late-peak (R2 = 0.200, P = .001) groups. Adjusted P is for the difference
after adjusting for BMI and total body fat mass

with lack of compensation of first- and second-phase insulin secretion
(first phase: 244.3 ± 26.7 vs 231.9 ± 26.7 μU/mL; second phase:
312.2 ± 31.2 vs 279.7 ± 20.7 μU/mL, Late-peak and Early-peak,
respectively). The best-fit hyperbolic relationship between insulin
sensitivity and first-phase insulin secretion showed a shift to the
left toward the origin in the Late-peak compared with the Early-
peak groups (Figure 3B). The significant differences in clamp-
measured parameters persisted after adjusting for BMI and total
body fat mass, with a borderline significance for peripheral insulin
sensitivity (Figures 2 and 3). The significant difference in hepatic
insulin sensitivity between the two groups remained even after
adjusting for glycemic status (ie, NGT vs IGT) in addition to BMI
and total body fat mass. This was not the case for peripheral insulin
sensitivity.
4 | DI SCU SSION
The present investigation of obese adolescents without type 2 diabe-
tes reveals that a Late-peak (glucose peak >30 minutes) vs Early-peak
(at 30 minutes) OGTT-glucose pattern harbors risk biomarkers of type
2 diabetes characterized by: (a) impaired β-cell function relative to
insulin sensitivity, (b) lower in vivo hepatic and peripheral insulin sen-
sitivity, (c) worse adipose tissue insulin resistance with higher FFA
concentrations despite higher prevailing insulinemia, (d) lower GIP
response during the OGTT with no differences in GLP-1 and glucagon,
and (e) higher prevalence of IGT.
Studies in adults demonstrated that delayed timing of the glucose
peak represents worse pathophysiological markers of type 2 diabetes
characterized by insulin resistance and lower DI in both non-diabetic
24
individuals8,13 and those with type 2 diabetes.17 These cross-sectional
observations were confirmed longitudinally in postpartum women,
showing that both a shift of the glucose peak to a later time point
(>30 minutes), and a consistently delayed glucose peak was associated
with declining β-cell function and worsening glucose tolerance over
time.16 The Relationship between Insulin Sensitivity and Cardiovascu-
lar Disease (RISC) study, using a novel latent class trajectory analysis
of heterogeneous OGTT glucose response patterns at baseline and
after 3 years of follow up, reported that time-to-glucose-peak was
almost exclusively determined by insulin sensitivity.8 Our findings in
pediatrics concur with the cross-sectional observations in adults, but
differ from the RISC results in that both insulin sensitivity and β-cell
function are impaired in obese adolescents with Late-peak vs Early-
peak glucose. The reasons behind this are likely multifactorial includ-
ing: (a) RISC used five different classes of time-to-glucose-peak while
we and all the other studies used two classes. Such detailed classifica-
tion, while possible with their large number of participants (n = 1566),
is not possible with our numbers and (b) different age groups, adults
31
in RISC and adolescents in ours. Recently our group and the Restor-
ing Insulin SEcretion (RISE) study32,33 showed that obese adolescents
with IGT or recent onset type 2 diabetes are 50% more insulin-
resistant compared with their obese adult counterparts. This worse
insulin resistance in youth may inflict a greater burden on a “vulnera-
ble” β-cell, hence our observation that both insulin sensitivity and
β-cell function are impaired in the Late-peak vs the Early-peak group.
Lastly, the BMI of our youth was remarkably higher (median 34.7 kg/
2 2
m ) than the median BMI of RISC adults (25.1 kg/m ). This higher BMI
in youth might impose a further accentuation in their insulin resis-
tance, resulting in an impaired balance between insulin sensitivity and
insulin secretion.
Compared to the wealth of publications in adults with respect to
the timing of the glucose peak, we identified only a single publication
in pediatrics.14 In that study, the OGTT of overweight/obese girls with
polycystic ovary syndrome (PCOS) showed an association between
the Late-peak of glucose and increased risk for type 2 diabetes
reflected in higher prevalence of prediabetes (combined IFG and IGT),
and lower oDI, derived from Matsuda index.14 However, contrary to
the OGTT data, their subgroup analysis of girls who underwent
hyperinsulinemic-euglycemic clamp showed no differences in tissue-
specific insulin sensitivity (adipose, hepatic, and peripheral) between
the Early-peak (n = 11) vs the Late-peak (n = 21) groups. This con-
trasts with our hyperinsulinemic-euglycemic clamp data in 102 non-
diabetic obese adolescents, exhibiting that adolescents with a glucose
peak >30 minutes during the OGTT have lower hepatic and peripheral
insulin sensitivity, without adequate compensation in first- and
second-phase insulin secretion, resulting in impaired β-cell function,
compared with those whose glucose peak occurred at 30 minutes.
This disparity in risk biomarkers between the Early-peak and Late-
peak groups was present despite similar adiposity parameters, includ-
ing BMI percentile, percent body fat and abdominal visceral adiposity,
and despite similar fasting glucose concentration and HbA1c. The
diminished DI or the β-cell function relative to insulin sensitivity in the
Late-peak group is considered the most powerful metabolic predictor
KIM ET AL.
for the future development of type 2 diabetes.34,35 Our OGTT-based
data in the total cohort were consistent with our clamp-based data
showing lower HOMA-IS, lower insulinogenic index and lower oDI in
the Late-peak vs the Early-peak groups. The discrepancy between our
findings and the only other pediatric study in the literature14 may
stem from differences in study population, only girls (54 with PCOS)
in the latter study vs obese boys and girls in ours (n = 278), diverse
ethnicity in the latter study (white, black, Hispanic, biracial) vs bal-
anced representation of whites and blacks in ours, and the number of
participants who had hyperinsulinemic-euglycemic clamp data across
the two studies. Additional studies in pediatrics are needed to confirm
or negate the scarce pediatric literature. More importantly, however,
with the recent emergence of the RISE data,32,33 any assumption that
a scientific phenomenon in youth will be similar to what is published
in adults, not warranting investigations in youth, will be doing the sci-
entific community a disfavor. Confirming or repudiating the existing
observations in adults as compared with youth is crucial.
Our current investigation expands knowledge in several novel
aspects including adipose insulin resistance, and incretin hormone dif-
ferences between the Early-peak and the Late-peak groups. Adoles-
cents in the Late-peak vs the Early-peak groups had higher FFA-AUC
despite higher insulin-AUC, suggestive of blunted regulation of lipoly-
sis by insulin and confirmatory of insulin resistance of lipid metabo-
lism.36 Given our previous observation that acute elevation of FFA
induces insulin resistance and deterioration of β-cell function relative
to insulin sensitivity in youth,37,38 it could be postulated that the insu-
lin resistance and the worse β-cell function in the Late-peak group
might be consequent to elevated FFA and lipotoxicity. Whether adi-
pose tissue insulin resistance is also present in adults with Late-peak
vs Early-peak glucose, or is unique to youth has not been
investigated.
Our finding of lower insulinogenic index during the OGTT in the
Late-peak vs the Early-peak groups, but similar first- and second-
phase insulin during the hyperglycemic clamp between the two
groups, supports the postulate that factors triggered by oral glucose
ingestion, such as defective secretion and/or action of incretin hor-
mones could play a role. In line with such a theory, our data showed
that youth in the Late-peak group have ~13% lower GIP secretion
during the OGTT compared with those in the Early-peak group, while
the only other pediatric study showed blunted GLP-1 secretion in
obese PCOS girls with glucose peak >30 minutes.14 Inconsistent find-
ings of GIP and GLP-1 responses in relation to the timing of the
glucose-peak between ours and the only other pediatric study14 might
be influenced by the fact that the number of participants with predia-
betes in the Late-peak group in our study (35%) was lower than that
of the others (52%). Differential enhancement and/or attenuation of
GIP vs GLP-1 secretion has been reported by varying degree of
weight status and glucose tolerance.21,39 In the present study, a lack
of significant difference in total GLP-1 between the two groups may
result in the absence of glucagon contribution to the OGTT-time-to-
glucose-peak as considering other non-insulinotropic effect of GLP-1
(ie, suppression of glucagon secretion). There was a randomized,
double-blind, placebo-controlled liraglutide trial in adults with type
KIM ET AL.
2 diabetes. The potential beneficial effect of liraglutide in shifting the
OGTT-time-to-glucose-peak, from later to an earlier peak, did not
40
reach statistical significance when compared to the placebo group.
Additional investigations are much needed to enhance knowledge
with respect to the relationship between incretin hormones and the
OGTT-time-to-glucose-peak. Moreover, given that separate contribu-
tions of GIP and GLP-1 to the potentiation of postprandial insulin
secretion have not been well understood,39 it would be crucial to
examine whether the incretin effect might be different between the
Early-peak vs Late-peak groups despite inconsistent results of incretin
concentrations.
Based on our data and the other pediatric study14 in non-diabetic
youth, the combination of insulin resistance in glucose and lipid
metabolism, β-cell dysfunction, and impaired incretin secretion could
explain the higher prevalence of IGT along with the higher 2-hour glu-
cose concentration, and the higher glucose-AUC and FFA-AUC in the
Late-peak vs the Early-peak groups that we observed. Similar findings
between the Late-peak vs the Early-peak groups, from either sub-
group analysis of NGT participants only or glycemic status adjustment
analysis in the total cohort, suggest that the time-to-glucose-peak
could be an early/sensitive indicator of heightened risk for prediabe-
tes and type 2 diabetes. However, to advance these cross-sectional
observations, prospective studies are needed to investigate the pre-
dictive value of the OGTT-time-to-glucose-peak for future develop-
ment of prediabetes or type 2 diabetes in obese at-risk youth. Adults
with early type 2 diabetes who underwent short-term intensive insu-
lin therapy for 4 to 6 weeks exhibited a shift to an earlier time-to-glu-
cose-peak together with improved β-cell function.15 Whether this
holds true in adolescents is questionable, especially that in the RISE
trial glargin for 3 months followed by metformin for 9 months or met-
formin alone was not effective in restoring or preserving β-cell func-
tion in youth.41
The strengths of the present investigation include: (a) the age of
the participants, with an in depth assessment of OGTT-time-to-glu-
cose-peak in adolescents with balanced inclusion of males and
females, and whites and blacks; (b) a first-time examination of
Adipose-IR in relation to the OGTT-time-to-glucose-peak; (c) the
novel and comprehensive assessment of incretin responses during the
OGTT in relation to OGTT-time-to-glucose-peak; and (d) the use of
the hyperinsulinemic-euglycemic clamp and the hyperglycemic clamp
to measure in vivo insulin sensitivity and insulin secretion relative to
OGTT-time-to-glucose-peak. Potential limitations would be that the
OGTT-time-to-glucose-peak was determined by a single OGTT, which
may have limited reproducibility in youth.42 Although a recent study
in adults showed that the OGTT-time-to-glucose-peak has good
agreement (73%) on three OGTTs,15 youth-specific investigation on
reproducibility of this marker is warranted. Another perceived but not
founded criticism could be that we previously published on the shape
of the glucose response curve in obese adolescents7 and that there
might be overlap. However, there is no evidence that the OGTT-glu-
cose-response-curve and the time-to-glucose-peak reflect similar
pathophysiological alterations. Chung et al13 examined the time-to-
glucose-peak and the glucose-response-curve simultaneously and
25
demonstrated that their predictive ability for prediabetes differed. In
their study, a Late-peak vs an Early-peak while associated with
impaired insulin sensitivity and secretion, with a 4-fold increased odds
of prediabetes, the monophasic curve vs biphasic did not show a sig-
nificant predictive ability for prediabetes.13
Our data in a large number of obese adolescents reveal that the
Late-peak (glucose peak occurring after 30 minutes) heralds risk bio-
markers of youth-onset type 2 diabetes, including poor β-cell function
relative to insulin sensitivity, impaired incretin secretion, global insulin
resistance in glucose and lipid metabolism, and higher prevalence of IGT.
AC KNOWLEDG EME NT S
The authors thank the children who participated in this study and
their parents; Nancy Guerra, C.R.N.P., for her assistance; Resa
Stauffer for her laboratory expertise; and the nursing staff of the Pedi-
atric Clinical and Translational Research Center for their outstanding
care of the participants and meticulous attention to the research. This
study was supported by grants from the National Institute of Child
Health and Human Development (K24-HD01357 and R01-HD27503
to S.A.), the American Diabetes Association (7-08-JF-27) and National
Institute of Diabetes and Digestive and Kidney Diseases grant
1R21DK083654-01A1 to S.L., National Center for Advancing Transla-
tional Sciences Clinical and Translational Science Award
UL1TR000005, and National Center for Research Resources grant
UL1RR024153 to the General Clinical Research Center.
CONFLIC T OF INT ER E ST
All authors have no conflicts of interest to declare.
AUTHOR CONTRIBU TIONS
J.Y.K. and S.A. designed the study, analyzed the data, and wrote the
manuscript. H.T., F.B., and S.L. contributed data and reviewed the
manuscript, S.F.M. collected and maintained the database and
reviewed the manuscript, and S.Y. and N.G. helped with data analysis
and reviewed the manuscript. All authors approved the manuscript in
its final version. S.A. is the guarantor of this work and, as such, had full
access to all the data in the study and takes responsibility for the
integrity of the data and the accuracy of the data analysis.
OR CID
Joon Young Kim https://orcid.org/0000-0003-0448-1684
SoJung Lee https://orcid.org/0000-0002-6634-6800
Sara F. Michaliszyn https://orcid.org/0000-0002-1176-9187
Silva Arslanian https://orcid.org/0000-0002-1528-9324
RE FE RE NCE S
1. American Diabetes Association. 2. Classification and diagnosis of dia-
betes: standards of medical care in diabetes 2018. Diabetes Care.
2018;41(suppl 1):S13-S27.
2. Unwin N, Shaw J, Zimmet P, Alberti KG. Impaired glucose tolerance
and impaired fasting glycaemia: the current status on definition and
intervention. Diabet Med. 2002;19(9):708-723.
3. Gerstein HC, Santaguida P, Raina P, et al. Annual incidence and rela-
tive risk of diabetes in people with various categories of dysglycemia:
26
a systematic overview and meta-analysis of prospective studies. Dia-
betes Res Clin Pract. 2007;78(3):305-312.
4. Tschritter O, Fritsche A, Shirkavand F, Machicao F, Haring H,
Stumvoll M. Assessing the shape of the glucose curve during an oral
glucose tolerance test. Diabetes Care. 2003;26(4):1026-1033.
5. Abdul-Ghani MA, Lyssenko V, Tuomi T, Defronzo RA, Groop L. The shape
of plasma glucose concentration curve during OGTT predicts future
risk of type 2 diabetes. Diabetes Metab Res Rev. 2010;26(4):280-286.
6. Kim JY, Coletta DK, Mandarino LJ, Shaibi GQ. Glucose response
curve and type 2 diabetes risk in Latino adolescents. Diabetes Care.
2012;35(9):1925-1930.
7. Kim JY, Michaliszyn SF, Nasr A, et al. The shape of the glucose
response curve during an oral glucose tolerance test heralds bio-
markers of type 2 diabetes risk in obese youth. Diabetes Care. 2016;
39(8):1431-1439.
8. Hulman A, Witte DR, Vistisen D, et al. Pathophysiological characteris-
tics underlying different glucose response curves: a latent class trajec-
tory analysis from the prospective EGIR-RISC study. Diabetes Care.
2018;41(8):1740-1748.
9. Abdul-Ghani MA, Abdul-Ghani T, Ali N, Defronzo RA. One-hour
plasma glucose concentration and the metabolic syndrome identify
subjects at high risk for future type 2 diabetes. Diabetes Care. 2008;
31(8):1650-1655.
10. Tfayli H, Lee SJ, Bacha F, Arslanian S. One-hour plasma glucose con-
centration during the OGTT: what does it tell about beta-cell function
relative to insulin sensitivity in overweight/obese children? Pediatr
Diabetes. 2011;12(6):572-579.
11. Kim JY, Goran MI, Toledo-Corral CM, Weigensberg MJ, Choi M,
Shaibi GQ. One-hour glucose during an oral glucose challenge pro-
spectively predicts beta-cell deterioration and prediabetes in obese
Hispanic youth. Diabetes Care. 2013;36(6):1681-1686.
12. Retnakaran R, Qi Y, Connelly PW, Sermer M, Hanley AJ, Zinman B.
Risk of early progression to prediabetes or diabetes in women with
recent gestational dysglycaemia but normal glucose tolerance at
3-month postpartum. Clin Endocrinol (Oxf). 2010;73(4):476-483.
13. Chung ST, Ha J, Onuzuruike AU, et al. Time to glucose peak during an
oral glucose tolerance test identifies prediabetes risk. Clin Endocrinol
(Oxf). 2017;87(5):484-491.
14. Cree-Green M, Xie D, Rahat H, et al. Oral glucose tolerance test glu-
cose peak time is most predictive of prediabetes and hepatic steatosis
in obese girls. J Endocr Soc. 2018;2(6):547-562.
15. Kramer CK, Vuksan V, Choi H, Zinman B, Retnakaran R. Emerging
parameters of the insulin and glucose response on the oral glucose
tolerance test: reproducibility and implications for glucose homeosta-
sis in individuals with and without diabetes. Diabetes Res Clin Pract.
2014;105(1):88-95.
16. Kramer CK, Ye C, Hanley AJ, et al. Delayed timing of post-challenge
peak blood glucose predicts declining beta cell function and worsen-
ing glucose tolerance over time: insight from the first year postpar-
tum. Diabetologia. 2015;58(6):1354-1362.
17. Wang X, Zhao X, Zhou R, et al. Delay in glucose peak time during the
oral glucose tolerance test as an indicator of insulin resistance and
insulin secretion in type 2 diabetes patients. J Diabetes Investig. 2018;
9(6):1288-1295.
18. Defronzo RA. Banting Lecture. From the triumvirate to the ominous
octet: a new paradigm for the treatment of type 2 diabetes mellitus.
Diabetes. 2009;58(4):773-795.
19. Kim JY, Nasr A, Tfayli H, Bacha F, Michaliszyn SF, Arslanian S. Increased
lipolysis, diminished adipose tissue insulin sensitivity, and impaired beta-
cell function relative to adipose tissue insulin sensitivity in obese youth
with impaired glucose tolerance. Diabetes. 2017;66(12):3085-3090.
20. Kim JY, Bacha F, Tfayli H, Michaliszyn SF, Yousuf S, Arslanian S. Adi-
pose tissue insulin resistance in youth on the spectrum from normal
weight to obese and from normal glucose tolerance to impaired glucose
tolerance to type 2 diabetes. Diabetes Care. 2019;42(2):265-272.
KIM ET AL.
21. Michaliszyn SF, Mari A, Lee S, et al. Beta-cell function, incretin effect,
and incretin hormones in obese youth along the span of glucose toler-
ance from normal to prediabetes to type 2 diabetes. Diabetes. 2014;
63(11):3846-3855.
22. Tfayli H, Lee S, Arslanian S. Declining beta-cell function relative to
insulin sensitivity with increasing fasting glucose levels in the
nondiabetic range in children. Diabetes Care. 2010;33(9):2024-2030.
23. Burns SF, Bacha F, Lee SJ, Tfayli H, Gungor N, Arslanian SA. Declining
beta-cell function relative to insulin sensitivity with escalating OGTT
2-h glucose concentrations in the nondiabetic through the diabetic
range in overweight youth. Diabetes Care. 2011;34(9):2033-2040.
24. Sjaarda LG, Bacha F, Lee S, Tfayli H, Andreatta E, Arslanian S. Oral dis-
position index in obese youth from normal to prediabetes to diabetes:
relationship to clamp disposition index. J Pediatr. 2012;161(1):51-57.
25. Sjaarda LA, Michaliszyn SF, Lee S, et al. HbA(1c) diagnostic categories
and beta-cell function relative to insulin sensitivity in over-
weight/obese adolescents. Diabetes Care. 2012;35(12):2559-2563.
26. Tanner JM. Growth and maturation during adolescence. Nutr Rev.
1981;39(2):43-55.
27. Lee S, Kuk JL, Kim Y, Arslanian SA. Measurement site of visceral adi-
pose tissue and prediction of metabolic syndrome in youth. Pediatr
Diabetes. 2011;12(3, pt 2):250-257.
28. Klopfenstein BJ, Kim MS, Krisky CM, Szumowski J, Rooney WD,
Purnell JQ. Comparison of 3 T MRI and CT for the measurement of
visceral and subcutaneous adipose tissue in humans. Br J Radiol.
2012;85(1018):e826-e830.
29. Bacha F, Gungor N, Lee S, Arslanian SA. In vivo insulin sensitivity and
secretion in obese youth: what are the differences between normal
glucose tolerance, impaired glucose tolerance, and type 2 diabetes?
Diabetes Care. 2009;32(1):100-105.
30. Abdul-Ghani MA, Molina-Carrion M, Jani R, Jenkinson C,
Defronzo RA. Adipocytes in subjects with impaired fasting glucose
and impaired glucose tolerance are resistant to the anti-lipolytic
effect of insulin. Acta Diabetol. 2008;45(3):147-150.
31. Arslanian S, Kim JY, Nasr A, et al. Insulin sensitivity across the lifespan
from obese adolescents to obese adults with impaired glucose toler-
ance: who is worse off? Pediatr Diabetes. 2018;19(2):205-211.
32. The RISE Consortium. Metabolic contrasts between youth and adults
with impaired glucose tolerance or recently diagnosed type 2 diabe-
tes: I. Observations using the hyperglycemic clamp. Diabetes Care.
2018;41(8):1696-1706.
33. The RISE Consortium. Metabolic contrasts between youth and adults
with impaired glucose tolerance or recently diagnosed type 2 diabe-
tes: II. Observations using the oral glucose tolerance test. Diabetes
Care. 2018;41(8):1707-1716.
34. Abdul-Ghani MA, Williams K, DeFronzo RA, Stern M. What is the
best predictor of future type 2 diabetes? Diabetes Care. 2007;30(6):
1544-1548.
35. Utzschneider KM, Prigeon RL, Faulenbach MV, et al. Oral disposi-
tion index predicts the development of future diabetes above and
beyond fasting and 2-h glucose levels. Diabetes Care. 2009;32(2):
335-341.
36. Periwal V, Chow CC, Bergman RN, Ricks M, Vega GL, Sumner AE.
Evaluation of quantitative models of the effect of insulin on lipolysis
and glucose disposal. Am J Physiol Regul Integr Comp Physiol. 2008;
295(4):R1089-R1096.
37. Hughan KS, Bonadonna RC, Lee S, Michaliszyn SF, Arslanian SA.
Beta-cell lipotoxicity after an overnight intravenous lipid challenge
and free fatty acid elevation in African American versus American
White overweight/obese adolescents. J Clin Endocrinol Metab. 2013;
98(5):2062-2069.
38. Michaliszyn SF, Bonadonna RC, Sjaarda LA, Lee S, Farchoukh L,
Arslanian SA. Beta-cell lipotoxicity in response to free fatty acid ele-
vation in prepubertal youth: African American versus Caucasian con-
trast. Diabetes. 2013;62(8):2917-2922.
KIM ET AL. 27

39. Chia CW, Egan JM. Incretins in obesity and diabetes. Ann N Y Acad SUPPORTING INF ORMATION
Sci. 2019;[Epub ahead of print]. Additional supporting information may be found online in the
40. Tran S, Kramer CK, Zinman B, Choi H, Retnakaran R. Effect of chronic
Supporting Information section at the end of this article.
liraglutide therapy and its withdrawal on time to postchallenge peak
glucose in type 2 diabetes. Am J Physiol Endocrinol Metab. 2018;314
(3):E287-E295.
41. The RISE Consortium. Impact of insulin and metformin versus metfor- How to cite this article: Kim JY, Tfayli H, Bacha F, et al. β-cell
min alone on beta-cell function in youth with impaired glucose toler- function, incretin response, and insulin sensitivity of glucose
ance or recently diagnosed type 2 diabetes. Diabetes Care. 2018;41 and fat metabolism in obese youth: Relationship to
(8):1717-1725.
OGTT-time-to-glucose-peak. Pediatr Diabetes. 2020;21:
42. Libman IM, Barinas-Mitchell E, Bartucci A, Robertson R, Arslanian S.
Reproducibility of the oral glucose tolerance test in overweight chil-
18–27. https://doi.org/10.1111/pedi.12940
dren. J Clin Endocrinol Metab. 2008;93(11):4231-4237.