University of Southern California Dissertations and Theses

Variation in insulin-like growth factor-2 binding protein 2 interacts with adiposity to alter insulin sensitivity in Mexican Americans

(USC Thesis Other)

Variation in insulin-like growth factor-2 binding protein 2 interacts with adiposity to alter insulin sensitivity in Mexican Americans

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

VARIATION IN INSULIN-LIKE GROWTH FACTOR-2 BINDING PROTEIN 2
INTERACTS WITH ADIPOSITY TO ALTER INSULIN SENSITIVITY IN MEXICAN
AMERICANS

by

Xia Li

________________________________________________________________________

A Thesis Presented to the
FALCULTY OF THE KECK SCHOOL OF MEDICINE
UNIVERSITY OF SOUTHERN CAILFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(APPLIED BIOSTATISTICS AND EPIDEMIOLOGY)

December 2008

Copyright 2008 Xia Li

ii
ACKNOWLEDGEMENTS
This paper would not have been possible without the guidance, instruction, and
encouragement of my thesis committee chair and advisor Dr. Richard Watanabe, under
whose supervision I chose this topic and began the thesis. This work was supported by
NIH grant DK-61628 and an American Diabetes Association Distinguished Clinical
Scientist Award to Dr. Thomas Buchanan. I would also like to thank Dr. Hooman
Allayee and Dr. Wendy Mack, my committee members in the final stages of the work,
who have also been very helpful and have assisted me in numerous ways, including
reviewing and giving suggestions on this paper. I cannot end without any gratitude for
my family, on whose constant encouragement and love I have relied throughout my time
at the University of Southern California.

iii
TABLE OF CONTENTS
Acknowledgments ii
List of Tables iv
List of Figures v
Abbreviation vi
Abstract viii
Introduction 1
Methods and Procedures 3
Subject Recruitment 3
Clinical Protocols 4
Assays 4
Molecular Analysis 4
Data Analysis 5
Results 8
Discussion 20
Bibliography 26

iv
LIST OF TABLES
Table 1. Subject characteristics 9
Table 2. Estimated haplotype frequencies 10
Table 3. Tag SNP characteristics 12
Table 4. Results of tests of association between SNPs in IGF2BP2 and T2DM-related
quantitative traits 14

Table 5. Results of interaction between rs11705701 with percent body fat on T2DM-
related quantitative traits 15

v
LIST OF FIGURES
Figure 1. IGF2BP2 pair-wise linkage disequlibrium and haplotype block structure 11
Figure 2. Interaction between rs11705701 and percent body fat on S
I
17
Figure 3. Interaction between rs11705701 and percent body fat on AIR 18
Figure 4. Interaction between rs11705701 and percent body fat on DI 19

vi
ABBREVIATIONS
IGF2BP2 Insulin-Like Growth Factor-2 Binding Protein 2
GWA Genome-Wide Association
T2DM Type 2 Diabetes Mellitus
GDM Gestational Diabetes Mellitus
IGF2 Insulin-Like Growth Factor 2
S
I
Insulin Sensitivity
AIR Acute Insulin Response
FUSION Finland-United States Investigation of Non-insulin Dependent Diabetes
Mellitus

IVGTT Intravenous Glucose Tolerance Test
OGTT Oral Glucose Tolerance Test
DEXA Dual Energy X-ray Absortiometry
SNP Single Nucleotide Polymorphism
LD Linkage Disequilibrium
ABI Applied Biosystems Incoorparation.
DI Disposition Index
HWE Hardy-Weinberg Equilibrium
SOLAR Sequential Oligogenic Linkage Analysis Routines
S
G
Glucose Effectiveness
BMI Body Mass Index

vii
MAF Minor Allele Frequency

df degree of freedom

AFBAC Affected Family-Based Controls test
TDT Transmission Disequilibrium Test
Q-TDT Quantitative trait Transmission Disequilibrium Test
FBAT Family-Based Association Test
PC Principle Components

viii
ABSTRACT
GWA studies have shown that variation in IGF2BP2 is associated with T2DM.
We examined a 20 kb region of IGF2BP2 for association with T2DM-related quantitative
traits in Mexican American families from the BetaGene study. We genotyped 14 SNPs
in 783 individuals from 146 families phenotyped by OGTT, IVGTT, and DEXA scan for
percent body fat. Three SNPs and one SNP combination were tested for association with
T2DM-related quantitative traits using a variance components framework. rs11705701
showed association with adiposity (P
ACT
=0.041), whereby body fat decreased ~1.5-2%
per copy of the A allele. We next tested whether rs11705701 interacted with body fat to
affect T2DM-related quantitative traits. The joint effect of rs11705701 and its interaction
with percent body fat was significantly associated with S
I
(Bonferroni p=0.028). We
conclude that rs11705701 is associated with body fat and this effect on body fat
influences insulin resistance which may contribute to T2DM.

1
INTRODUCTION
GWA studies in Caucasian samples identified insulin-like growth factor-2 binding
protein 2 (IGF2BP2) as a susceptibility gene for T2DM (Saxena, Voight et al. 2007;
Scott, Mohlke et al. 2007; Zeggini, Weedon et al. 2007). IGF2BP2 is part of a family of
binding proteins (Jones and Clemmons 1995; Nielsen, Christiansen et al. 1999).
IGF2BP1 binds to the leader 3 mRNA in the 5’-UTR of the insulin-like growth factor 2
(IGF2) gene to regulate IGF2 translation (Nielsen, Christiansen et al. 1999). Mobility
shift experiments show that both IGF2BP2 and IGF2BP3 bind to a specific RNA segment
with similar affinity to that of IGF2BP1, suggesting that all three members of this binding
protein family may regulate IGF2 translation in a similar manner (Nielsen, Christiansen
et al. 1999). IGF2, in turn, is a polypeptide growth factor that plays important roles in
growth and development and stimulates insulin action. However, the exact functional
role of IGF2BP2 is not known and even less is known about how variation in IGF2BP2
may affect quantitative traits that contribute to diabetes risk.
We have shown that Mexican American women with a previous diagnosis of
GDM have a 5-year T2DM risk that exceeds 50% (Kjos, Peters et al. 1995; Peters, Kjos
et al. 1996), an observation that has now been extended to other populations (Kim,
Newton et al. 2002). We have also shown that the increased risk for T2DM in these
women is due, in part, to a progressive loss of β-cell function in the face of chronic
insulin resistance (Buchanan, Metzger et al. 1990; Buchanan, Xiang et al. 1998;
Buchanan, Xiang et al. 1999). Furthermore, risk for T2DM in these women can be

2
significantly reduced by ameliorating insulin resistance, thereby reducing insulin
secretory demand on the pancreatic β-cells and improving β-cell function (Buchanan,
Xiang et al. 2002; Xiang, Peters et al. 2006). Therefore, the study of women with
previous GDM provides an opportunity to examine potential genetic influences on
T2DM-related quantitative traits.
The BetaGene Study is a family-based study to identify such influences where we
have hypothesized that if there exists genetic variation influencing T2DM-related
quantitative traits, then such variants should be segregating within families. In BetaGene,
we are performing detailed phenotyping of Mexican American probands with recent
GDM and their family members to obtain quantitative estimates of body composition, S
I
,
AIR, and β-cell compensation (DI), which are traits associated with the pathogenesis of
T2DM (cf. (DeFronzo 1988; Bergman 1989; Reaven 1993)) and shown to be heritable in
Mexican American families (Bergman, Zaccaro et al. 2003; Watanabe and Ouzunian
2003). The present report focuses on the effects of genetic variation in a 20 kb region
surrounding rs1470579 in IGF2BP2, based on unpublished preliminary results from the
stage 1 analysis of the FUSION Study GWA (Scott, Mohlke et al. 2007), which
suggested that variation in this region may be associated with S
I
derived from an IVGTT.
We determined that rs11705701, a variant that appears to be correlated with an
alternative splice form of IGF2BP2 (Prokunina-Olsson, Swift et al. 2008), is associated
with percent body fat and that the interaction between this variant and adiposity is
associated with S
I
.

3
METHODS AND PROCEDURES
Subject Recruitment
Subject recruitment for BetaGene is on-going and for the purpose of this report
we describe only those subjects, clinical protocols and assays related to the results
presented herein. Subjects are Mexican American (both parents and ≥3 grandparents
Mexican or of Mexican descent). Self-reported parental and grandparental birthplace
information was obtained to assess ancestry. Subjects are either probands with GDM
diagnosed within the previous five years and their family members, or non-GDM
probands with normal glucose levels in pregnancy in the past five years. All probands
are identified from the patient populations at Los Angeles County/USC Medical Center,
the Kaiser Permanente of Southern California health plan membership, and OB/GYN
clinics at local hospitals. GDM probands are recruited for phenotyping if they, (a) have a
confirmed diagnosis of GDM within the previous five years, (b) have glucose levels
associated with poor pancreatic β-cell function and a high risk of diabetes when not
pregnant (Kjos, Peters et al. 1995), (c) have no evidence of β-cell autoimmunity by GAD-
65 testing, and (d) have available for study at least two siblings. If at least two first
cousins of the GDM proband are available from the same nuclear family, those cousins
are recruited as well. Non-GDM probands are recruited if they had a 1-hour 50 g glucose
screening result <130 mg/dl (7.2 mM) during their most recent pregnancy and have no
family history of diabetes. They are frequency-matched in equal ratios to GDM probands
by age, BMI, and parity. Subjects with chronic medical conditions or chronic medication
use known to have important effects on glucose metabolism are excluded.

4
All protocols for BetaGene have been approved by the Institutional Review
Boards of participating institutions and all participants provided written informed consent
prior to participation.
Clinical Protocols
Phenotyping is performed on two separate visits to the General Clinical Research
Center. Visit 1 includes a study-directed medical history and physical examination,
screening serum chemistries and complete blood count, DNA collection, and a 75 g
OGTT with blood samples obtained before and 30, 60, 90, and 120 min after glucose
ingestion. Participants from GDM families with fasting glucose <126 mg/dl on the
OGTT and non-GDM probands with normal OGTT results are invited for a second visit,
which consists of a DEXA scan for determination of body composition and an IVGTT
performed as previously described (Buchanan, Xiang et al. 1999).
Assays
Plasma glucose is measured on an auto-analyzer using the glucose oxidase
method (YSI Model 2300, Yellow Springs Instruments, Yellow Springs, OH). Insulin is
measured by two-site immunoenzymometric assay (TOSOH) that has <0.1% cross-
reactivity with proinsulin and intermediate split products.
Molecular Analysis
Given the unpublished preliminary results from the FUSION study, we focused
our examination of IGF2BP2 to a 20 Kb region surrounding rs1470579. Because it was
equivocal whether tag SNP selection based upon HapMap data would reflect the
underlying LD pattern in Mexican Americans, we selected SNPs at ~2.5 Kb intervals

5
across this 20 Kb region, preferentially selecting SNPs that were polymorphic in all four
HapMap populations (HapMap Build 35). Fourteen SNPs, including rs4402960, the SNP
originally shown to be associated with T2DM in Caucasians (Saxena, Voight et al. 2007;
Scott, Mohlke et al. 2007; Zeggini, Weedon et al. 2007), were selected and genotyped
using the Applied Biosystems Incorporation (ABI) TaqMan system (Livak 1999; Livak
2003). These SNPs were all located within introns 1 or 2 of IGF2BP2. Genotyping
assays were either selected through ABI’s “Assays on Demand” database
(http://myscience.appliedbiosystems.com/navigation/mysciapplications.jsp) or custom
designed using ABI’s “Assays by Design” service.
Data Analysis
We calculated two measures of insulin response to glucose: the difference
between the 30’ and fasting plasma insulin concentration from OGTT (30’ Δinsulin) and
the incremental area under the insulin curve for during the first 10 min of the IVGTT
(AIR). IVGTT glucose and insulin data were analyzed using the minimal model
(MINMOD Millennium V5.18) to derive measures of S
G
and S
I
(Bergman, Ider et al.
1979; Boston, Stefanovski et al. 2003). Disposition Index (DI) was computed as the
product of S
I
and AIR and used as a measure of β-cell compensation for insulin resistance
(Bergman, Phillips et al. 1981).
Genotype data were tested for deviation from Hardy-Weinberg Equilibrium
(HWE) and non-Mendelian inheritance using PEDCHECK V1.1 (O'Connell and Weeks
1998) and PEDSTATS V0.6.4 (Wigginton and Abecasis 2005). Allele frequencies were
estimated by maximum likelihood using data from all available family members and

6
SOLAR V2.1.4 (Blangero and Almasy 1997; Almasy and Blangero 1998). Haplotype
frequencies were estimated using FUGUE (www.sph.umich.edu/csg/abecasis/FUGUE/).
Tag SNPs were selected from among the genotyped SNPs using TAGGER with 2- and 3-
maker haplotypes as implemented in Haploview V3.32 (Barrett, Fry et al. 2005). In
addition, LD and haplotype block structure were assessed using Haploview and the
method of Gabriel (Gabriel, Schaffner et al. 2002).
Quantitative trait data (BMI, % body fat, fasting glucose, 2-hour glucose, fasting
insulin, 30’ ∆insulin, 2-hour insulin, S
G
, S
I
, AIR, and DI) were statistically transformed to
approximate univariate normality using the Box-Cox power transform method in SAS 9.1
prior to analyses. (Box 1964). Association between SNPs and quantitative traits was
tested using a 1-df likelihood ratio test (Equation 1 and 2) under a variance components
framework as implemented in SOLAR (Blangero and Almasy 1997; Almasy and
Blangero 1998). Each SNP was initially tested for the association with quantitative traits
assuming an additive genetic model (SNP was coded 0, 1, and 2, as the number of minor
alleles). The AG combination of rs13060777 and rs6444082 (cf. Results) was modeled
as the number of copies of the AG haplotype of these two SNPs. Models were adjusted
for continuous age and gender (Equation 1), or adjusted for continous age, gender, and
percent body fat (Equation 2). Because our families were ascertained through probands

7
based on previous GDM status, to extend our findings to the general population, we
corrected for ascertainment bias by conditioning each model on the proband’s GDM
phenotype. Any result under the additive model that showed a trend for an association
between SNPs and T2DM-related quantitative traits, defined as p=0.1 with Bonferroni
correction for the number of markers examined (number of markers=4, corrected
p=0.025), was subsequently tested for association under dominant (SNP was coded 0 and
1, representing the presence of any minor allele) and/or recessive (SNP was coded 0 and
1, representing homozygosity of the minor allele) genetic models. Since Bonferroni
correction for multiple testing is overly conservative for correlated tests, we used P
ACT
to
better control for the number of tests performed and to account for the residual
correlation among the tag SNPs and additional correlation among the quantitative traits
tested for association (Conneely and Boehnke 2007).
The univariate association between rs11705701 and percent body fat (cf. Results)
led us to examine whether the association between rs11705701 and T2DM-related
quantitative traits was modified by percent body fat by testing the interaction between
rs11705701 and body fat. We compared a model with age, gender, and percent body fat
as main effects with a second model that included a main effect for rs11705701 and the
multiplicative interaction between rs11705701 and percent body fat. We tested for the
overall gene effect using a 2-df test (Equation 3) on all T2DM-related quantitative traits
(fasting glucose, 2-hour glucose, fasting insulin, 30’ ∆insulin, 2-hour insulin, S
G
, S
I
, AIR,

8
and DI). P
ACT
is not applicable under the 2-df test (Conneely and Boehnke 2007), so for
these analyses, we performed a Bonferroni correction to account for the multiple traits
tested. However, because many of the quantitative traits examined are correlated and
Bonferroni assumes independence, the corrected p-values can be considered be overly
conservative. Any result under the additive model that showed a trend for overall genetic
association, defined as p=0.1 Bonferroni-corrected for the number of traits examined
(number of traits=9, corrected p=0.011), was subsequently tested for a significant
interaction effect using a 1-df test under the additive model (Equation 4).
Linear modeling results are reported as age, gender, and percent body fat-
adjusted means and standard deviations. Results from the interaction analysis that
included percent body fat are shown in figures, and trait values are reported as age- and
gender-adjusted means and standard deviations. All other results were reported as
unadjusted medians and interquartile ranges.

RESULTS
We report results from 783 individuals in 146 families for whom all phenotype
and genotype data were available (Table 1). There was an average of 5.0 (range 1-11)
members from each family including probands, siblings, and cousins. GDM probands,
siblings, cousins, and non-GDM probands were of similar age, although cousins were

9
Table 1. Subject Characteristics.

Values are unadjusted median and interquartile range.

GDM Probands Siblings Cousins Non-GDM Probands
Female/Male 118/0 229/150 125/95 66/0
Age [yrs.] 35.8 (5.8) 35.1 (8.7) 33.0 (9.2)* ** 34.6 (5.0)
BMI [Kg/m
2
] 32.5 (7.0) 29.9 (6.1)* 28.6 (6.4)* ** 28.8 (6.1)* ** ***
Body Fat [%] 39.2 (5.3) 32.8 (8.5)* 30.9 (8.6)* ** 36.9 (5.7)* ** ***
Fasting Glucose [mM] 5.4 (0.6) 5.2 (0.5)* 5.1 (0.6)* 4.9 (0.4)* ** ***
2-hour Glucose [mM] 9.2 (2.3) 7.7 (2.2)* 6.9 (2.0)* ** 6.3 (1.2)* ** ***
Fasting Insulin [pM] 83 (57) 62 (45)* 59 (52)* 47 (35)* ** ***
30’ ∆Insulin [pM] 417 (287) 463 (325) 490 (342)* 453 (322)
2-hour Insulin [pM] 721 (500) 526 (411)* 451 (385)* ** 353 (255)* ** ***
S
G
[ ×10
-2
min
-1
] 1.48 (0.48) 1.71 (0.68)* 1.83 (0.71)* ** 2.01 (0.69)* **
S
I
[ ×10
-3
min
-1
per pM] 2.76 (1.76) 2.94 (1.63) 3.16 (1.73)* 3.35 (1.46)*
AIR [pM × 10 min] 4171 (4281) 5678 (4521)* 6245 (5285)* 7096 (5109)* **
Disposition Index 8694 (5829) 14055 (9579)* 15710 (9394)* ** 19333 (9294)* ** ***
* uncorrected p < 0.05 vs. probands
** uncorrected p < 0.05 vs. siblings
*** uncorrected p < 0.05 vs. Cousins

10
significantly younger compared to siblings and probands (p=0.0067). Among probands,
siblings, and cousins, median BMI exceeded the threshold for being considered
overweight (25 Kg/m
2
). Median BMI and percent body fat were both significantly lower
in siblings (p=0.01 for both), cousins (p<0.0001 for both) and non-GDM probands
(p<0.0001 for both) compared to GDM probands. Overall, quantitative trait values for
glucose regulation were higher in non-GDM probands, siblings, and cousins compared to
GDM probands (Table 1).
Among the 14 SNPs genotyped, 2 SNPs (rs1374910, rs9862583) failed the test for
HWE, and 3 SNPs (rs9843367, rs9869293, and rs9867882) had MAF <1% in our sample
and were not further analyzed. The average inter-maker distance among the 9 remaining
SNPs was 2.97 Kb with the smallest gap being 1.2 Kb and the largest 8.1 Kb. Estimated
pair-wise LD and haplotype blocks are shown in Figure 1. Eight of the 9 SNPs formed a
single 24 Kb haplotype block comprised of 5 haplotypes with frequency >1% (Table 2).
The observed LD and haplotype block patterns were similar, although not identical, to the

Table 2. Estimated Haplotype Frequencies.

rs4402960
rs6767484
rs13060777
rs11711477
rs1470579
rs6769511
rs9859406
rs6111082
Frequency
G A A T A T G G 0.729
T G G A C C A T 0.130
T G A A C C A T 0.084
T G G A C C A G 0.027
T G A A C C A G 0.016

11
Figure 1. IGF2BP2 Pair-wise Linkage Disequlibrium and Haplotype Block
Structure.

Panel A shows pair-wise LD and haplotype block structure determined by the Gabriel method as
implemented in Haploview V4.0 for SNPs genotyped in the BetaGene study. Tagging SNPs are
identified by a box. A single 24 kb haplotype block was identified and rs11705701 only was not
included in this block. Panel B shows pair-wise LD and haplotype block structure for the identical
SNPs from the HapMap CEPH (Utah residents with ancestry from northern and western Europe)
sample. A similar LD and haplotype block pattern was observed for the combined HapMap
Japanese in Tokyo and Han Chinese in Beijing samples. LD is displayed as pair-wise r2 values
(values within boxes), where white indicate r2 = 0, varying shades of grey indicate 0 < r2 < 1, and
black indicates r2 = 1.

12
Table 3. Tag SNP Characteristics.

Tag SNP
Position
[Kb]*
Minor
Allele
MAF SNPs Tagged
rs13060777 187007901 G 0.15 rs13060777
rs6444082 187018925 T 0.23 rs6444082
rs11705701 187027011 A 0.33 rs11705701
rs1306077/rs6444082 --- AG ---
rs4402960, rs11711477, rs6769511,
rs1470579, rs9859406, rs6767484
* NCBI Build 35

pattern derived from the HapMap CEPH samples (Figure 1B), where there is a single 32
kb haplotype block encompassing all SNPs. A similar LD and haplotype block pattern
was observed in the combined HapMap Japanese and Han Chinese samples (data not
shown), while the Yoruban samples showed multiple smaller haplotype blocks across this
region (data not shown). Tagger identified 3 SNPs (rs13060777, rs6444082, and
rs11705701) and one SNP pair (AG combination of rs13060777 and rs6444082), as tags
for this region of IGF2BP2 (Table 3). The AG combination captured both rs1470579,
which was the original SNP showing evidence of association with S
I
in the FUSION
Study GWA (Scott, Mohlke et al. 2007), and rs4402960, which was initially shown to be
associated with T2MD in Caucasians (Saxena, Voight et al. 2007; Scott, Mohlke et al.
2007; Zeggini, Weedon et al. 2007). Thus these three tag SNPs and one SNP pair that
captured the majority of the variation in the region were selected to test for association
with T2DM-related quantitative traits.
Two of the tag SNPs (rs6444082, rs11705701) and the AG combination were
nominally associated with percent body fat under an additive genetic model (Table 4,

13
nominal p=0.016, 0.001, and 0.011 respectively). These same SNPs also showed
nominal association with 30’ Δinsulin and 2-hour insulin from the OGTT (Table 4).
Despite the nominal association of rs64440872, rs11705701, and the AG combination
with percent body fat, none showed evidence for association with BMI (p>0.12, cf.
Table4). For results showing a trend for association under the additive genetic model,
neither dominant nor recessive genetic models showed stronger associations, suggesting
the association is best characterized by an additive model (data not shown).
The most significant association between rs11705701 and percent body fat
remained significant even after P
ACT
was applied to account for the correlation among
SNPs and phenotypes and for multiple testing (PACT=0.041). Percent body fat decrease
~1.5-2% (corresponding to ~2.5-4 pounds of fat) with each copy of the A allele for
rs11705701 after adjusted for age and gender (GG=33.1±6.1%, GA=31.5±6.3%,
AA=29.5±6.6%). Given the association with body fat, we next assessed whether adding
percent body fat as an additional covariate altered our results regarding genetic
associations for other T2DM-related quantitative traits. All nominal associations initially
observed when only age and gender were included as covariates became non-significant
when percent body fat was added as a covariate to the analysis (Table 4).
The modest association between rs11705701 and percent body fat and the strong
effect that percent body fat has on T2DM-related quantitative traits suggested the percent
body fat and rs11705701 could interact to affect these metabolic variables. To assess this
possibility, we initially tested whether there was an overall association between T2DM-
related quantitative traits and the joint effect of rs11705701 and its interaction with

14
Table 4. Results of Tests of Association between SNPs in IGF2BP2 and T2DM-related Quantitative Traits.

Uncorrected p-values are shown for test of association under an additive genetic model adjusted for age and gender or with the
additional adjustment for percent body fat.

rs13060777 rs6444082 rs11705701 AG Combination**
Effect
*
Uncorrected
p-value

Effect
*
Uncorrected
p-value

Effect
*
Uncorrected
p-value

Effect
*
Uncorrected
p-value

Age,
Gender
Age,
Gender,
% Fat

Age,
Gender
Age,
Gender,
% Fat

Age,
Gender
Age,
Gender,
% Fat

Age,
Gender
Age,
Gender,
% Fat
BMI - 0.783 N/A - 0.124 N/A - 0.309 N/A + 0.244 N/A
% Body Fat - 0.134 N/A - 0.016 N/A - 0.001 N/A + 0.011 N/A
Fasting Glucose - 0.422 0.736 - 0.677 0.885 - 0.720 0.664 + 0.645 0.813
2-hour Glucose + 0.963 0.850 + 0.386 0.942 + 0.225 0.834 - 0.418 0.862
Fasting Insulin - 0.530 0.940 - 0.186 0.774 + 0.352 0.479 + 0.386 0.483
30’ ∆Insulin + 0.125 0.311 + 0.040 0.137 + 0.018 0.109 - 0.035 0.141
2-hour Insulin - 0.055 0.104 - 0.014 0.175 - 0.008 0.126 + 0.011 0.156
S
G
+ 0.660 0.850 + 0.926 0.790 - 0.256 0.100 - 0.922 0.722
S
I
+ 0.718 0.595 + 0.090 0.967 + 0.150 0.743 - 0.166 0.703
AIR - 0.199 0.396 - 0.051 0.249 - 0.047 0.191 + 0.093 0.337
DI - 0.154 0.135 - 0.340 0.182 - 0.200 0.065 + 0.361 0.169
* Refers to the directional effect of the minor allele on the quantitative trait based on the regression coefficient.
** The AG combination of rs13060777 and rs6444082 was modeled as the number of copies of the AG haplotype (see Methods
and Procedures for details).

15
Table 5. Results of Interaction between rs11705701 with Percent Body Fat on
T2DM-related Quantitative Traits.

Uncorrected p-values are shown for test of association under an additive genetic model adjusted
for age and gender.

2 degree of freedom test* 1 degree of freedom test**

Uncorrected
p value
Corrected
p value***

Uncorrected
p value
Corrected
p value****
Fasting Glucose 0.613 1 N/A N/A
2-hour Glucose 0.093 0.837 N/A N/A
Fasting Insulin 0.256 1 N/A N/A
30’ ∆Insulin 0.275 1 N/A N/A
2-hour Insulin 0.007 0.063 0.006 0.019
S
G
0.107 0.963 N/A N/A
S
I
0.003 0.027 0.001 0.002
AIR 0.418 1 N/A N/A
DI 0.008 0.072 0.012 0.037
* The joint effect of rs11705701 and its interaction with percent body fat using a 2-df test.
** The interaction between rs11705701 and percent body fat alone using a 1-df test.
*** p values are Bonferroni corrected by the number of traits (n=9)
**** p values are Bonferroni corrected by the number of traits (n=3)

percent body fat using a 2-df test (Table 5). The joint effect was marginally associated
with OGTT 2-hour insulin (Bonferroni p=0.066) and DI (Bonferroni p=0.072) and
significantly associated with S
I
(Bonferroni p=0.028). The interaction between
rs11705701 and percent body fat alone was also significantly associated with these traits
(Table 5): OGTT 2-hour insulin (Bonferroni interaction p=0.019), S
I
(Bonferroni
interaction p=0.002), and DI (Bonferroni interaction p=0.037).
Figure 2A (top) illustrates the association between rs11705701 and S
I
stratified by
percent body fat tertile. Figure 2B (bottom) illustrates the same effect using the model-

16
predicted trait values to better visualize the interaction between rs11705701 and percent
body fat. At low body fat, S
I
was highest in AA homozygotes and lowest in GG
homozygotes. S
I
decreased with increasing percent body fat across all three genotype
groups, illustrating the presumed effect of adiposity to induce insulin resistance.
However, the decline in S
I
across the range of percent body fat followed the opposite
order of the pattern observed at low body fat. The fall in S
I
was greatest among AA
homozygotes, less for heterozygotes, and the least for GG homozygotes.
By comparison, we did not observe a significant interaction between rs11705701
and adiposity on AIR (Figure 3). By itself, rs11705701 was only marginally associated
with AIR (uncorrected p=0.047). AIR tended to increase with increasing body fat, but
was reduced with each copy of the A allele across the range of percent body fat observed
in our sample.
Simultaneous examination of Figures 2 and 3 reveals an apparent dissociation
between insulin sensitivity and normal β-cell compensation. With increased adiposity, S
I

fell most in AA individuals, while manifesting the smallest rise in AIR across the range
of body fat. This lack of compensatory insulin secretion hints at a possible β-cell defect,
which is reflected in Figure 4. Figure 4 demonstrates an interaction between rs11705701
and percent body fat on DI, a measure of β-cell compensation that is derived from the
IVGTT. DI decreased with increasing percent body fat in all genotype groups, but the
decline was most precipitous among AA homozygous individuals. In the lowest body fat
tertile, all three genotypes had similar DI, but in the middle and highest body fat tertile,

17
Figure 2. Interaction between rs11705701 and Percent Body Fat on S
I
.

Panel A shows age and gender-adjusted mean S
I
and standard deviation stratified by rs11705701
genotype assuming an additive genetic model and percent body fat tertiles. There was an overall
effect of rs11705701 on S
I
(Bonferroni 2-df p=0.028) and the interaction between rs11705701
and percent body fat was significantly associated with S
I
(Bonferroni p=0.002). S
I
decreases with
increasing body fat among all rs11705701 genotypes; however, the decrease is least among GG
homozygotes and greatest among AA homozygotes. Panel B shows the interaction based on the
model parameter estimates and covering the range of percent body fat observed in the BetaGene
study. The consequence of the differential effect of the variant across the range of body fat is
clearly depicted where in the low body fat range S
I
is higher with each copy of the A allele, while
within the high body fat range S
I
is lower with each additional copy of the A allele.
G/G G/A A/A
0
1
2
3
4
5
Low Medium High
Covariate-adjusted S
I
[ ×10
-3
×min
-1
per pmol/L]
rs11705701 Genotype
10 20 30 40 50 60
0
2
4
6
8
10
G/G G/A A/A
Predicted Covariate-adjusted S
I
[ ×10
-3
×min
-1
per pmol/L]
% Body Fat

18
Figure 3. Interaction between rs11705701 and Percent Body Fat on AIR.

Panel A shows age and gender-adjusted mean AIR and standard deviation stratified by
rs11705701 genotype assuming an additive genetic model and percent body fat tertiles.
rs11705701 was marginally associated with AIR (uncorrected p=0.047) and there was no
evidence for percent body fat to modify this association. AIR is similar within each genotype
group, indicating no effect of percent body fat to alter AIR. Panel B shows the interaction based
on the model parameter estimates and covering the range of percent body fat observed in the
BetaGene study. AIR increases with increasing body fat similarly for all genotype groups across
the range of body fat. AIR is lower with each copy of the A allele across the range of body fat.
G/G G/A A/A
0
2
4
6
8
Low Medium High
Covariate-adjusted AIR
[ ×10
3
pmol/L ×10 min]
rs11705701 Genotype
10 20 30 40 50 60
0
2
4
6
8
10
G/G G/A A/A
Predicted Covariate-adjusted AIR
[ ×10
3
pmol/L ×10 min]
% Body Fat

19
Figure 4. Interaction between rs11705701 and Percent Body Fat on DI.

Panel A shows age and gender-adjusted mean DI and standard deviation stratified by rs11705701
genotype assuming an additive genetic model and percent body fat tertiles. There was a marginal
effect of rs11705701 on DI (Bonferroni 2-df p=0.072) and the interaction between rs11705701
and percent body fat was significantly associated with DI (Bonferroni p=0.037). Within each
genotype group, DI is lower as body fat increases, but the fall in DI is higher among AA
homozygotes compared to GG homozygotes. Panel B shows the interaction based on the model
parameter estimates and covering the range of percent body fat observed in the BetaGene study.
The consequence of the differential effect of the variant across the range of body fat is clearly
depicted where in the low body fat range DI is higher with each copy of the A allele, while within
the high body fat range DI is lower with each additional copy of the A allele.

G/G G/A A/A
0
5
10
15
20
Low Medium High
Covariate-adjusted DI
rs11705701 Genotype
10 20 30 40 50 60
0
5
10
15
20
G/G G/A A/A
Predicted Covariate-adjusted DI
% Body Fat

20
DI among AA homozygotes was lowest among the three genotype groups. Thus, the
deleterious effect of adiposity on β-cell function was greatest among individuals with the
AA genotype.

DISCUSSION
The association of SNP rs4402960, which is located in intron 2 of IGF2BP2, was
initially shown to be associated with T2DM in Caucasian samples (Saxena, Voight et al.
2007; Scott, Mohlke et al. 2007; Zeggini, Weedon et al. 2007) and has been replicated in
several other populations (Grarup, Rose et al. 2007; Horikoshi, Hara et al. 2007; Pascoe,
Tura et al. 2007; Omori, Tanaka et al. 2008). However, to our knowledge evidence for
association of IGF2BP2 with T2DM-related quantitative traits has not been reported.
One group reported a modest association between rs4402960 and insulin resistance as
determined by the HOMA-index in a Japanese sample (Horikoshi, Hara et al. 2007) while
the IRAS Family study reported an association between rs4402960 and DI in their
Hispanic sample (Palmer, Goodarzi et al. 2008). Our approach differed from these prior
studies in three ways. First, we used more sophisticated and direct methods for
phenotyping body composition, insulin resistance, and β-cell function. Second, we
applied those methods to a Mexican American population with a wide range of diabetes
risk. Third, instead of focusing specifically on rs4402960 or SNPs in high LD with it, we
screened a 20 kb region of IGF2BP2, including rs4402960, to test for association with
our carefully ascertained traits. With this approach, we observed nominal association

21
between the AG combination of rs13060777 and rs6444082, which tags rs4402960, and
percent body fat, OGTT 30’ Δinsulin, and OGTT 2-hour insulin. However, these did not
remain significant after correction for multiple testing. By contrast, we found significant
associations between a separate SNP in the region, rs11705701, and several T2DM-
related quantitative traits. SNPs rs11705701 and rs4402960 are ~32.6 kb apart and in
strong LD (D'=0.902), although they are only modestly correlated (r2=0.579) in our
sample. rs11705701 is in the promoter region ~1.48 kb upstream from exon 1 of
IGF2BP2 and is not included in the haplotype block that encompasses all other SNPs we
genotyped in our sample (cf. Figure 1). The fact that our associations are observed with
rs11705701 and not rs4402960 suggests the former may be in LD with a SNP with
functional consequences for IGF2BP2 in Mexican Americans.
The most significant association that we observed was between rs11705701 and
percent body fat, which yielded a nominal p-value of 0.001, and remained significant
after correction for multiple testing and the correlation among both SNPs and traits
(P
ACT
=0.041). This finding led us to test whether associations between rs11705701 and
other T2DM-related quantitative traits were modified by body fat. Without testing the
effect of body fat, we observed only a nominal association between rs11705701 and
OGTT 2-hour insulin (uncorrected p=0.008), which did not remain significant after
correction for multiple testing, and did not observe associations between rs11705701 and
either S
I
or DI (uncorrected p=0.15, 0.2, respectively, Table 4). Taking into
consideration the interaction between body fat, rs11705701, and other T2DM-related
traits, we found that body fat has very strong effects on the association between

22
rs11705701 and S
I
and modest effects on OGTT 2-hour insulin and DI. Our results
showed an overall effect of rs11705701 on these traits, as reflected by the 2-df test of the
SNP main effect and its interaction with percent body fat, while the significant interaction
term in the 1-df test points to heterogeneity with respect to percent body fat (Table 5).
Plots of the data revealed that both S
I
and DI decreased most with increasing body fat in
AA homozygotes but decreased least for GG homozygotes. Of note, there were no
analogous associations with BMI, suggesting that IGF2BP2 has its main effects on body
fat and demonstrating the potential limitation of BMI as a surrogate measure of body fat
in the context of genetic studies.
Population stratification occurs when allele frequencies are different across
subgroups of the population, which also have different baseline risks of the disease,
thereby leading to false-positive associations in studies (Knowler, Williams et al. 1988;
Thomas and Witte 2002). Since Mexican Americans are likely a mixture of non-
Hispanic white, Native American, and Latino American heritage and we used self-
reported data, it is possible that population stratification may confound our analysis.
Family based data can be used in association studies to avoid potential ethnic
mismatching between cases and randomly determined controls; therefore, several family-
based association tests were developed as appropriate strategies to control for population
stratification. For example, the AFBAC can be used to test the association between a
genetic marker and a single locus model of disease in nuclear families which have at least
one affected child. In order to avoid population stratification, this test designed a
matched case-control study, where the cases were parental marker alleles transmitted

23
from parents to an affected child and the controls were alleles not transmitted (Thomson
1995). In addition, the original TDT was designed to test for both LD and linkage in
parents-offspring trios that contain at least one affected offspring by measuring allelic
transmission from a heterozygous parent to the affected child (Spielman, McGinnis et al.
1993). The Q-TDT is an extension of the TDT developed for quantitative traits (Allison
1997). By comparison, the FBAT can be used with dichotomous, quantitative, and
censored traits and missing parental genotype information by processing
transmission/disequilibrium type tests as a two-stage procedure. In the first stage, the test
statistics for the association between the trait locus and the marker locus are calculated.
In the second stage, false positive rates are corrected by comparing the test statistics to
conditional distribution of the genetic model given the minimal sufficient statistics under
the null hypothesis (Laird, Horvath et al. 2000; Rabinowitz and Laird 2000).
Furthermore, PC analysis that is based on the EIGENSTRAT method can be used to find
individual components of population substructure using PCs which are weighted by
corresponding eigenvectors. Since the PCs are independent to the population structure,
by adding the first few PCs with the largest eigenvalues as covariables in the regression
model, we can correct population stratification in the association studies between genetic
markers and quantitative traits in population-based data (Price, Patterson et al. 2006).
In this study, we tested the association between SNPs and quantitative traits using
a likelihood ratio test under a variance components framework as implemented in
SOLAR instead of the tests mentioned above. We did not incorporate PC analysis
because of the limited number of markers in our study, which makes accurate assessment

24
of population substructure difficult. In addition, two main limitations exist in family-
based tests that prevented their application. First, tests based on allelic transmissions,
such as AFBAC, Q-TDT, and FBAC, will reduce the efficiency of the tests relative to the
variance components approach. Second, some of these tests, such as AFBAC and TDT,
can only be applied to dichotomous traits. Therefore, the approach we used herein can
(a) be applied for testing association between genetic marker and quantitative data in
family-based data, (b) control for ascertainment bias, and (c) provide higher efficiency
for association tests compared to family-based tests, although it cannot control for
population stratification. Nonetheless, we cite two lines of evidence to suggest that
population stratification is not an issue in this study. First, our results remain similar
when an individual’s birthplace is included as an additional covariate in our analyses.
While birthplace is not a perfect proxy for ancestry, the fact that birthplace has little
effect on our association results would suggest that ancestry is not a major contributor to
our observations. Another reason is that when we re-examine the association between
rs11705701 and percent body fat using the FBAT (Laird, Horvath et al. 2000; Rabinowitz
and Laird 2000), which is protected from population stratification, we still see evidence
for the association (p=0.018). The reduction in the magnitude of the association is
expected, given the reasons outlined above.
In summary, we tested variation in a limited 20 kb region of IGF2BP2 for
association with T2DM-related quantitative traits in our sample of Mexican Americans in
the BetaGene study. Our selection of the limited 20 kb region was motivated by
preliminary results emanating from the FUSION study. We show that rs11705701 is

25
associated with percent body fat in our subjects and may be associated with insulin
secretion. We also show that S
I
is associated with the interaction between rs11705701
and percent body fat and that β-cell function is marginally associated with the same
interaction. Our results are supported by in vitro studies from the FUSION study
showing that rs11705701 correlates with an alternative splice form of IGF2BP2 that
appears to be primarily expressed in adipose tissue and pancreatic islets. We conclude
that variation in IGF2BP2 is associated with body fat. This effect on body fat influences
insulin resistance in peripheral tissues and in combination with a modest effect of
IGF2BP2 on insulin secretion, results in poor β-cell function, which in turn contributes to
risk for type 2 diabetes.

26
BIBLIOGRAPHY
Allison, D. B. (1997). "Transmission-disequilibrium tests for quantitative traits." Am J
Hum Genet 60(3): 676-90.
Almasy, L. and J. Blangero (1998). "Multipoint quantitative-trait linkage analysis in
general pedigrees." Am J Hum Genet 62(5): 1198-211.
Barrett, J. C., B. Fry, et al. (2005). "Haploview: analysis and visualization of LD and
haplotype maps." Bioinformatics 21(2): 263-5.
Bergman, R. N. (1989). "Lilly lecture 1989. Toward physiological understanding of
glucose tolerance. Minimal-model approach." Diabetes 38(12): 1512-27.
Bergman, R. N., Y. Z. Ider, et al. (1979). "Quantitative estimation of insulin sensitivity."
Am J Physiol 236(6): E667-77.
Bergman, R. N., L. S. Phillips, et al. (1981). "Physiologic evaluation of factors
controlling glucose tolerance in man: measurement of insulin sensitivity and beta-cell
glucose sensitivity from the response to intravenous glucose." J Clin Invest 68(6): 1456-
67.
Bergman, R. N., D. J. Zaccaro, et al. (2003). "Minimal model-based insulin sensitivity
has greater heritability and a different genetic basis than homeostasis model assessment
or fasting insulin." Diabetes 52(8): 2168-74.
Blangero, J. and L. Almasy (1997). "Multipoint oligogenic linkage analysis of
quantitative traits." Genet Epidemiol 14(6): 959-64.
Boston, R. C., D. Stefanovski, et al. (2003). "MINMOD Millennium: a computer
program to calculate glucose effectiveness and insulin sensitivity from the frequently
sampled intravenous glucose tolerance test." Diabetes Technol Ther 5(6): 1003-15.
Box, G. E. P., Cox, D.R. (1964). "An analysis of transformations." Journal of the Royal
Statistical Society 26: 211-52.

27
Buchanan, T. A., B. E. Metzger, et al. (1990). "Insulin sensitivity and B-cell
responsiveness to glucose during late pregnancy in lean and moderately obese women
with normal glucose tolerance or mild gestational diabetes." Am J Obstet Gynecol
162(4): 1008-14.
Buchanan, T. A., A. Xiang, et al. (1998). "Gestational diabetes: antepartum
characteristics that predict postpartum glucose intolerance and type 2 diabetes in Latino
women." Diabetes 47(8): 1302-10.
Buchanan, T. A., A. H. Xiang, et al. (1999). "Antepartum predictors of the development
of type 2 diabetes in Latino women 11-26 months after pregnancies complicated by
gestational diabetes." Diabetes 48(12): 2430-6.
Buchanan, T. A., A. H. Xiang, et al. (2002). "Preservation of pancreatic beta-cell function
and prevention of type 2 diabetes by pharmacological treatment of insulin resistance in
high-risk hispanic women." Diabetes 51(9): 2796-803.
Conneely, K. N. and M. Boehnke (2007). "So Many Correlated Tests, So Little Time!
Rapid Adjustment of P Values for Multiple Correlated Tests." Am J Hum Genet 81(6).
DeFronzo, R. A. (1988). "Lilly lecture 1987. The triumvirate: beta-cell, muscle, liver. A
collusion responsible for NIDDM." Diabetes 37(6): 667-87.
Gabriel, S. B., S. F. Schaffner, et al. (2002). "The structure of haplotype blocks in the
human genome." Science 296(5576): 2225-9.
Grarup, N., C. S. Rose, et al. (2007). "Studies of association of variants near the HHEX,
CDKN2A/B, and IGF2BP2 genes with type 2 diabetes and impaired insulin release in
10,705 Danish subjects: validation and extension of genome-wide association studies."
Diabetes 56(12): 3105-11.
Horikoshi, M., K. Hara, et al. (2007). "Variations in the HHEX gene are associated with
increased risk of type 2 diabetes in the Japanese population." Diabetologia 50(12): 2461-
6.
Jones, J. I. and D. R. Clemmons (1995). "Insulin-like growth factors and their binding
proteins: biological actions." Endocr Rev 16(1): 3-34.

28
Kim, C., K. M. Newton, et al. (2002). "Gestational diabetes and the incidence of type 2
diabetes: a systematic review." Diabetes Care 25(10): 1862-8.
Kjos, S. L., R. K. Peters, et al. (1995). "Predicting future diabetes in Latino women with
gestational diabetes. Utility of early postpartum glucose tolerance testing." Diabetes
44(5): 586-91.
Knowler, W. C., R. C. Williams, et al. (1988). "Gm3;5,13,14 and type 2 diabetes
mellitus: an association in American Indians with genetic admixture." Am J Hum Genet
43(4): 520-6.
Laird, N. M., S. Horvath, et al. (2000). "Implementing a unified approach to family-based
tests of association." Genet Epidemiol 19 Suppl 1: S36-42.
Livak, K. J. (1999). "Allelic discrimination using fluorogenic probes and the 5' nuclease
assay." Genet Anal 14(5-6): 143-9.
Livak, K. J. (2003). "SNP genotyping by the 5'-nuclease reaction." Methods Mol Biol
212: 129-47.
Nielsen, J., J. Christiansen, et al. (1999). "A family of insulin-like growth factor II
mRNA-binding proteins represses translation in late development." Mol Cell Biol 19(2):
1262-70.
O'Connell, J. R. and D. E. Weeks (1998). "PedCheck: a program for identification of
genotype incompatibilities in linkage analysis." Am J Hum Genet 63(1): 259-66.
Omori, S., Y. Tanaka, et al. (2008). "Association of CDKAL1, IGF2BP2, CDKN2A/B,
HHEX, SLC30A8, and KCNJ11 with susceptibility to type 2 diabetes in a Japanese
population." Diabetes 57(3): 791-5.
Palmer, N. D., M. O. Goodarzi, et al. (2008). "Quantitative trait analysis of type 2
diabetes susceptibility loci identified from whole genome association studies in the
Insulin Resistance Atherosclerosis Family Study." Diabetes 57(4): 1093-100.

29
Pascoe, L., A. Tura, et al. (2007). "Common variants of the novel type 2 diabetes genes
CDKAL1 and HHEX/IDE are associated with decreased pancreatic beta-cell function."
Diabetes 56(12): 3101-4.
Peters, R. K., S. L. Kjos, et al. (1996). "Long-term diabetogenic effect of single
pregnancy in women with previous gestational diabetes mellitus." Lancet 347(8996):
227-30.
Price, A. L., N. J. Patterson, et al. (2006). "Principal components analysis corrects for
stratification in genome-wide association studies." Nat Genet 38(8): 904-9.
Prokunina-Olsson, L., Hanisch,JJ, Jackson,AU, Chines,PS, Erdos,MR, Bonnycastle,LL,,
A. Swift, Narisu,N, Scott,LJ, Morken,MA, Buchanan,TA, Mohlke,KL, Tuomilehto,J,, et
al. (2008). "A novel variant of IGF2BP2 is associated with type 2 diabetes and affects
expression of a functional splicing isoform." Diabetes 57 (Suppl 1): A61.
Rabinowitz, D. and N. Laird (2000). "A unified approach to adjusting association tests
for population admixture with arbitrary pedigree structure and arbitrary missing marker
information." Hum Hered 50(4): 211-23.
Reaven, G. M. (1993). "Role of insulin resistance in human disease (syndrome X): an
expanded definition." Annu Rev Med 44: 121-31.
Saxena, R., B. F. Voight, et al. (2007). "Genome-wide association analysis identifies loci
for type 2 diabetes and triglyceride levels." Science 316(5829): 1331-6.
Scott, L. J., K. L. Mohlke, et al. (2007). "A genome-wide association study of type 2
diabetes in Finns detects multiple susceptibility variants." Science 316(5829): 1341-5.
Spielman, R. S., R. E. McGinnis, et al. (1993). "Transmission test for linkage
disequilibrium: the insulin gene region and insulin-dependent diabetes mellitus (IDDM)."
Am J Hum Genet 52(3): 506-16.
Thomas, D. C. and J. S. Witte (2002). "Point: population stratification: a problem for
case-control studies of candidate-gene associations?" Cancer Epidemiol Biomarkers Prev
11(6): 505-12.

30
Thomson, G. (1995). "Mapping disease genes: family-based association studies." Am J
Hum Genet 57(2): 487-98.
Watanabe, R., Xiang,A, Trigo,E, Hernandez,S, Berrios,F, Langefeld,C, Kjos,S, and J.
Ouzunian, Sacks,D, Lawrence,J, Buchanan,T (2003). "Evidence of genetic predisposition
for Bcell dysfunction in Mexican-American families of probands with gestational
diabetes." Diabetes 52.
Wigginton, J. E. and G. R. Abecasis (2005). "PEDSTATS: descriptive statistics, graphics
and quality assessment for gene mapping data." Bioinformatics 21(16): 3445-7.
Xiang, A. H., R. K. Peters, et al. (2006). "Effect of pioglitazone on pancreatic beta-cell
function and diabetes risk in Hispanic women with prior gestational diabetes." Diabetes
55(2): 517-22.
Zeggini, E., M. N. Weedon, et al. (2007). "Replication of genome-wide association
signals in UK samples reveals risk loci for type 2 diabetes." Science 316(5829): 1336-41.

Abstract (if available)

Abstract GWA studies have shown that variation in IGF2BP2 is associated with T2DM. We examined a 20 kb region of IGF2BP2 for association with T2DM-related quantitative traits in Mexican American families from the BetaGene study. We genotyped 14 SNPs in 783 individuals from 146 families phenotyped by OGTT, IVGTT, and DEXA scan for percent body fat. Three SNPs and one SNP combination were tested for association with T2DM-related quantitative traits using a variance components framework. rs11705701 showed association with adiposity (P_ACT=0.041), whereby body fat decreased ~1.5-2% per copy of the A allele. We next tested whether rs11705701 interacted with body fat to affect T2DM-related quantitative traits. The joint effect of rs11705701 and its interaction with percent body fat was significantly associated with S_I (Bonferroni p=0.028). We conclude that rs11705701 is associated with body fat and this effect on body fat influences insulin resistance which may contribute to T2DM.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Variation in CRY2 and MTNR1B have independent effects on insulin secretion in Mexican Americans

PDF

Association of single nucleotide polymorphisms in GCK, GCKR and PNPLA3 with type 2 diabetes related quantitative traits in Mexican-American population

PDF

Ectopic fat and adipose tissue inflammation in overweight and obese African Americans and Hispanics

PDF

Adipokines do not account for the association between osteocalcin and insulin sensitivity in Mexican Americans

PDF

Quantity versus quality: how adipose tissue accumulation and immune cell profile associate with risk for type 2 diabetes in minority children and adults

PDF

The association between the PNPLA3 rs738409 variant and hepatic fat in Mexican Americans

Asset Metadata

Creator Li, Xia (author)

Core Title Variation in insulin-like growth factor-2 binding protein 2 interacts with adiposity to alter insulin sensitivity in Mexican Americans

School Keck School of Medicine

Degree Master of Science

Degree Program Applied Biostatistics

Publication Date 11/14/2008

Defense Date 11/02/2008

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag adiposity,genetics,IGF2BP2,insulin resistance,Mexican American,OAI-PMH Harvest,type 2 diabetes

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Watanabe, Richard M. (committee chair), Allayee, Hooman (committee member), Mack, Wendy J. (committee member)

Creator Email ionalee@usc.edu,xial@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m1766

Unique identifier UC1269083

Identifier etd-Li-2522 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-97796 (legacy record id),usctheses-m1766 (legacy record id)

Legacy Identifier etd-Li-2522.pdf

Dmrecord 97796

Document Type Thesis

Rights Li, Xia

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu