A Genome-Wide Association Study Confirms Previously Reported Loci for Type 2 Diabetes in Han Chinese
et al. (2011) A Genome-Wide Association Study Confirms Previously Reported Loci for Type 2 Diabetes in Han
Chinese. PLoS ONE 6(7): e22353. doi:10.1371/journal.pone.0022353
A Genome-Wide Association Study Confirms Previously Reported Loci for Type 2 Diabetes in Han Chinese
Bin Cui 0
Xiaolin Zhu 0
Min Xu 0
Ting Guo 0
Dalong Zhu 0
Gang Chen 0
Xuejun Li 0
Lingyan Xu 0
Yufang Bi 0
Yuhong Chen 0
Yu Xu 0
Xiaoying Li 0
Weiqing Wang 0
Haifeng Wang 0
Wei Huang 0
Kerby Shedden, University of Michigan, United States of America
0 1 Key Laboratory for Endocrine and Metabolic Diseases of Ministry of Health, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine , Shanghai , China , 2 Department of Endocrine and Metabolic Diseases, Shanghai Clinical Center for Endocrine and Metabolic Diseases, Shanghai Institute of Endocrine and Metabolic Diseases, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine , Shanghai , China , 3 Department of Endocrine Diseases, Nanjing Drum Tower Hospital, Affiliated Hospital of Nanjing University Medical School , Nanjing , China , 4 Department of Endocrine Diseases, Fujian Provincial Hospital , Fuzhou , China , 5 Department of Endocrine Diseases, First Affiliated Hospital of Xiamen University , Xiamen , China , 6 Department of Genetics, Chinese National Human Genome Center , Shanghai , China
Background: Genome-wide association study (GWAS) has identified more than 30 loci associated with type 2 diabetes (T2D) in Caucasians. However, genomic understanding of T2D in Asians, especially Han Chinese, is still limited. Methods and Principal Findings: A two-stage GWAS was performed in Han Chinese from Mainland China. The discovery stage included 793 T2D cases and 806 healthy controls genotyped using Illumina Human 660- and 610-Quad BeadChips; and the replication stage included two independent case-control populations (a total of 4445 T2D cases and 4458 controls) genotyped using TaqMan assay. We validated the associations of KCNQ1 (rs163182, p = 2.085610217, OR 1.28) and C2CD4A/ B (rs1370176, p = 3.67761024, OR 1.124; rs1436953, p = 7.75361026, OR 1.141; rs7172432, p = 4.00161025, OR 1.134) in Han Chinese. Conclusions and Significance: Our study represents the first GWAS of T2D with both discovery and replication sample sets recruited from Han Chinese men and women residing in Mainland China. We confirmed the associations of KCNQ1 and C2CD4A/B with T2D, with the latter for the first time being examined in Han Chinese. Arguably, eight more independent loci were replicated in our GWAS.
Funding: This study was supported by the grants from Key Laboratory for Endocrine and Metabolic Diseases of Ministry of Chinese Health (No. 1994DP131044),
State Key Laboratory of Medical Genomics, China, National Key New Drug Creation and Manufacturing Program (2008ZX09312/019) and the Sector Funds of
Ministry of Health (201002002). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
. These authors contributed equally to this work.
Type 2 diabetes (T2D) is a complex disease hallmarked by
insulin resistance and pancreatic beta-cell dysfunction . T2D is
becoming a major concern of global public health due to its
escalating prevalence throughout the world . In China, 9.7%
and 15.5% of the entire population suffer from T2D and
prediabetes, respectively . Although overfeeding and sedentary
lifestyle are claimed as the main contributors to its increasing
incidence, genetic factors play a significant role in the etiology and
pathogenesis of T2D, in many cases via interaction with
environmental counterparts .
Genetic analysis of T2D and related diseases (such as obesity
and monogenic diabetes) and traits (such as fasting plasma glucose
levels) has greatly improved our understanding of glucose
homeostasis and energy balance in both physiological and
pathological conditions, some of which has brought novel
preventive and therapeutic options . Before the year 2007,
studies based on linkage analysis and candidate gene strategy
identified only a few genetic loci of T2D . More recently,
several genome-wide association studies (GWASs) have been
completed in independent population samples derived from
Caucasians and Japanese, and identified a host of novel
susceptibility loci of T2D . With a more recent large-scale
meta-analysis taking into account , these studies in total have
discovered at least 38 independent susceptibility loci of T2D,
many of which have been replicated in populations of different
ancestries, including Han Chinese [21,22].
Extending GWAS to populations of diverse descents is
valuable, because different frequencies of genetic variants and
patterns of linkage disequilibrium (LD) due to population
backgrounds may strongly affect the power and potential of
GWAS to discover and/or refine certain genetic loci associated
with disease . For example, association of KCNQ1 with T2D
was first independently identified by two GWASs in Japanese
[14,15], although for both studies the sample size of the discovery
stage was small and the genomic coverage of SNPs insufficiently
dense . However, to date, original data of GWAS of T2D in
Han Chinese is still limited [25,26]. One study was conducted
among Han Chinese in Taiwan that identified two additional
novel loci in the protein tyrosine phosphatase receptor type D
(PTPRD) and serine racemase (SRR) genes . The other was
performed in women from Shanghai Womens Health Study
(SWHS) and Shanghai Breast Cancer Study (SBCS) . Here,
we reported a 2-stage study, comprising one discovery and one
replication stage, in both Han Chinese men and women residing
in Mainland China.
In the discovery stage, 998 T2D patients of Han Chinese
descent recruited in Shanghai were genotyped using Illumina
Human 660-Quad BeadChips. Our control population was
derived from a glucose survey in a community in Shanghai and
was genotyped using Illumina Human 660- and 610-Quad
BeadChips. After stringent quality control, we obtained 561,694
SNPs that were common to both genotyping platforms with both
average call rates of .99%. To ensure that case and control
groups were genetically matched, in addition to close examination
of their geographic origins, MDS was used to exclude population
outliers, the result of which was further confirmed by PCA, which
showed minimal evidence for population stratification [Figure S1,
Figure S2]. 474,515 SNPs in 793 cases and 806 controls entered
final statistical analysis using the Cochran-Armitage trend test to
examine the genotype-phenotype association under an additive
model. After genomic control (GC) with an inflation factor l of
1.08 [Figure S3], the association results did not change
We selected the top 30 significantly associated SNPs
representing 15 genomic regions in the discovery stage (arbitrarily defined
as trend-P #1024) for genotyping in an independent case-control
population from Shanghai (n = 2620) (1058 cases and 1562
controls, part of Replication 1) as a fast-track replication analysis
[Table S1, Figure S4]. Among the 30 SNPs, 3 representing 2
genomic loci were replicated with the same direction of association
with the discovery stage at the significance level of P,0.10
(Figure 1). Two of them, namely rs163182 and rs163184
(P = 2.34861027, OR = 1.37, and P = 0.04332, OR = 1.121,
respectively), were located in the same locus in gene KCNQ1 as
previously reported. The remaining SNP, rs3773159 (P = 0.09591,
OR = 1.152), is likely to represent a novel susceptibility locus for
T2D in the population.
We genotyped the rest of replication 1 (1602 cases and 506
controls residing in Shanghai and Jiangsu Province) and pooled
analysis (Replication 1: 2660 cases and 2068 controls) showed a
nominally significant association of rs3773159 with T2D
(P = 0.0418, OR = 1.137, Table 1). Further genotyping and
analysis in an independent sample comprising 1785 cases and
2390 controls (Replication 2: Southern Han) identified a consistent
association though without reaching significance level. Therefore,
there was not enough evidence to establish the association of
rs3773159 with T2D in out populations.
We checked a total of 37 genomic loci previously reported to be
associated with T2D  in our discovery stage data, among which
42 SNPs in 26 loci were successfully genotyped. 13 SNPs representing
9 loci were significantly associated with T2D in our GWAS at a
significant level of P,0.05 (Table S2): rs2793831 (proxy for
rs10923931, NOTCH2), rs2943641 (IRS1), rs2120825 (proxy for
rs1801282, PPARG), rs734312 (WFS1), rs896854 (TP53INP1),
rs10906115 (CDC123/CAMK1D), rs7901695 (TCF7L2), rs7903146
(TCF7L2), rs1552224 (CENTD2), rs1370176 (C2CD4A/B), rs1436953
(C2CD4A/B), rs7172432 (C2CD4A/B), and rs1436955 (C2CD4B). If
we adopt a less stringent threshold of P,0.10, 4 more SNPs
representing 3 genomic regions showed significant association with
T2D in the discovery stage (Table S2): rs1111875 (HHEX),
rs7923837 (HHEX), rs8050136 (FTO), and rs780094 (GCKR).
We noticed that the three SNPs rs1370176, rs1436953, and
rs7172432 in C2CD4A/B on chromosome 15 were exactly the
lead SNPs in a three-stage GWAS recently reported involving a
total of &19000 Japanese . Further genotyping in the
fasttrack replication sample showed that the three SNPs tended to be
consistently associated with T2D (P = 0.1134, 0.03651, and
0.01884, respectively). Genotyping in the remaining 1602 cases
and 506 controls of Replication 1 and pooled analysis (Replication
1) expectedly rendered these associations significant (rs1370176,
P = 0.03597, OR = 1.115; rs1436953, P = 9.67161025, OR =
1.187; rs7172432, P = 4.52961024, OR = 1.187, respectively;
Table 1). Further analysis in the Southern Han population
(Replication 2) showed that the associations were in the same
direction as those in Central Han population but did not reach
significance level (Table 1, Figure 1).
Our GWAS did not provide sufficient evidence for potentially
novel genetic variation associated with T2D in Han Chinese.
However, our study validated 10 previously reported loci
associated with T2D, including KCNQ1 and C2CD4A/B, and
several of which were very recently discovered by large-scale
multistage study and meta-analyses.
In Han Chinese population, our study validated previously
known susceptibility loci of T2D, among which SNPs in
KCNQ1were consistently and most strongly associated with T2D
(rs163182, Pcombined = 2.085610217, OR = 1.280). rs163184, the
SNP most significantly associated with T2D in KCNQ1 in our
discovery stage, is also the lead SNP in two other independent
studies [14,20]. One additional SNP in the same locus, namely
rs2237892, was reported to be significantly associated with T2D
in two independent studies in Japanese [14,15], and replicated
in our previous study . Collectively, these data confirmed
that variants in KCNQ1 were associated with T2D in different
populations, and our GWAS worked in identifying such
Aside from variants in KCNQ1, our study rediscovered several
SNPs in additional susceptibility loci of T2D originally identified
in populations of different ancestries, and SNPs in three such loci
came to our notice. The first is C2CD4A/B in 15q21.3, which was
quite recently reported by a GWAS of T2D involving about 8000
Japanese in their discovery stage . The most significant SNP
reported, rs7172432, was also most strongly associated with T2D
at the locus in the Han Chinese population (PGWAS = 3.26461024),
but did not reach the cutoff P = 1024, mainly due to our limited
power because of small sample size. Two more SNPs reported,
namely rs1370176 and rs1436953, were in the same locus and
associated with T2D in our GWAS. We noticed that another
recent work reported an association of rs1436955 with T2D in
Han Chinese, which was indeed in the same locus; albeit in silico
replication strategy might have prevented their further analysis
. Although not replicated in Southern Han population possibly
due to population reasons, the 3 SNPs were significantly associated
with T2D in Central Han population, with risk alleles having
slightly stronger effect sizes (OR = 1.1151.187) than those in
Japanese population as previously reported . These results
present direct evidence that genetic variants in C2CD4A/B locus
are associated with T2D in Central Han Chinese population
residing in Mainland China. Further studies are required to
investigate whether such associations exist in Southern and
Northern Han Chinese populations and identify the causal
The other two loci are TP53INP1 (rs896854, PGWAS = 0.002212)
and CENTD2 (rs1552224, PGWAS = 0.04226); both of them were
discovered in a large-scale meta-analysis comprising more than
one hundred thousand individuals of European descent
(DIAGRAM+). Our GWAS was likely to validate these two newly
identified loci in the Han Chinese population, but further
replication is required to examine their effects. Moreover, in light
of the small sample size of our GWAS stage, these results support
the notion that GWAS in Han Chinese has potential to identify
novel risk loci of T2D.
Our study represents the first GWAS with both discovery and
replication sample sets recruited both Han Chinese men and
women residing in Mainland China. Han Chinese is
geographically and genetically heterogeneous and has subpopulation
structures, which may have considerable effect on design and
interpretation of GWAS . Allele effects in Southern Han
Chinese were consistently weaker than those in Central Han
Chinese (Table 1); we consider this as a result of different
population backgrounds. Because T2D-associated variants show
much weaker effects than alleles associated with other diseases and
traits (e.g., autoimmune disease and malignancies), a much larger
RAF(T2D) and RAF(NC), risk allele frequency in T2D cases and controls, respectively. OR, odds ratio for risk allele.
sample consisting of homogenous individuals is required for
genuine associations to achieve genome-wide significance. Though
our initial sample size is small, which might have prevented us to
discover more potentially novel associations, our GWAS has
successfully replicated 10 previously reported T2D susceptibility
loci, several of which are very recently discovered by large-scale
studies and meta-analyses in Caucasians and Japanese. This fact
lends convincing evidence of the soundness of our study, and
highlights the potential of discovering novel genetic variations
associated with T2D by extending GWAS to diverse populations.
In conclusion, our genome-wide association study confirmed
several T2D susceptibility loci previously identified in Caucasians
and Japanese, among which variants in KCNQ1showed the
strongest association, and variants in C2CD4A/B were first
replicated in Han Chinese residing in Mainland China.
This study was approved by the Institutional Review Board of
the Ruijin Hospital, Shanghai Jiao Tong University School of
Medicine and was in accordance with the principle of the Helsinki
Declaration II. The written informed consent was obtained from
The GWAS genotyped 998 T2D patients recruited from
outpatient departments of Ruijin Hospital, and 1002 healthy
controls obtained from one glucose survey in Youyi community,
Baoshan district, Shanghai . The replication analysis included
two independent populations (Replication 1: 2660 T2D cases and
2068 controls residing in Shanghai  and Jiangsu province
[Central Han]; Replication 2: 1785 T2D cases and 2390 controls
residing in Southern China [Southern Han]) . All participants
self-reported as Han Chinese. T2D case was diagnosed according
to the 1999 World Health Organization criteria (fasting plasma
glucose level $7.0 mmol/l and/or 2h oral glucose tolerance test
(OGTT) plasma glucose level $11.1 mmol/l) or with taking
antidiabetic therapies. The controls were defined as a fasting glucose
level less than 6.1 mmol/l and a 2 h OGTT plasma glucose level
less than 7.8 mmol/l.
Genotyping and quality control
Genomic DNA was extracted from peripheral blood by
standard phenol/chloroform-based method. In the discovery
stage, genotyping was conducted by using Illumina Human
660and 610-Quad BeadChips at the Chinese National Human
Genome Center at Shanghai. Genotyping was performed
according to the Infinium HD protocol from Illumina. Quality
control involved exclusion of SNPs with a call rate ,90%, a
minimum allele frequency ,0.01, and/or a significant deviation
from Hardy-Weinberg equilibrium (HWE) in the controls
(P,1027). SNPs on the X, Y and mitochondrial chromosomes
and copy number variation (CNV) probes were also excluded from
In the replication stage, genotyping was conducted using 59
nuclease allelic discrimination assay (TaqMan Assay) on an ABI
PRISM 7900HT Sequence Detection System following the
manufacturers protocol. In our study, the call rate ranged from
97.95% (rs3773159 in the Replication 1) to 99.33% (rs3773159 in
Replication 2). There is no significant difference of SNP calling
between the case and the control groups. The average consensus
rate in the duplicate samples (n = 100) was 100%.
Identification of cryptic relatedness among individuals in the
discovery stage was based on pairwise identity by state using the
PLINK 1.07 software , after which one of the two related
individuals was excluded. Population structure of the remaining
sample was examined and outliers were excluded based on
multidimensional scaling analysis (MDS) using PLINK 1.07 as well
as principal component analysis (PCA) using EIGENSTRAT
software . Cochran-Armitage trend test was used to examine
the association of genotype with disease phenotype and calculate
odds ratio (OR) per allele. The quantile-quantile plot was
employed to evaluate the overall significance of the genome-wide
association results and impacts of population stratification. The
genomic control inflation factor lwas also calculated to examine
the effects of population stratification.
Replication analysis was performed by first analyzing
replication samples separately and then combining them with the
discovery sample set . Association analysis of the combined
samples was performed based on Cochran-Mantel-Haenszel tests
. Joint analysis was performed using PLINK under a
fixedeffect model .
Figure S1 Multidimensional scaling analysis (MDS)
plot. MDS plot by PLINK of the 793 cases and 806 controls
shows no evident population stratification and outliers. Blue:
control; pink: case.
Figure S2 Principal component analysis (PCA) plot. PCA
plot by EIGENSTRAT of the 793 cases and 806 controls shows
no evident population stratification and outliers. Green: control;
Figure S3 Quantile-quantile (Q-Q) plot for the trend
test. (l = 1.08). Q-Q plot for the Cochran-Armitage trend test
for 474,515 SNPs in 793 cases and 806 controls. l = 1.08 and
minimal evidence of association due to population stratification
Figure S4 Manhattan plot for GWAS data. The x-axis
represents chromosomal location of 474,515 SNPs examined and
the y-axis represents log10 of the P value of the
CochranArmitage trend test under an additive model. A cutoff line was
drawn at the significance threshold of 1024.
Table S1 SNPs selected for fast-track replication.
RAF(T2D) and RAF(NC), risk allele frequency in T2D cases
and controls, respectively. OR, odds ratio for risk allele.
Table S2 SNPs in previously reported T2D loci.
RAF(T2D) and RAF(NC), risk allele frequency in T2D cases
and controls, respectively. OR, odds ratio for risk allele. *, P,0.05.
The authors thank the field workers for their contribution and the
participants for their cooperation, as well as Minglan Yang, Liying Zhu,
Yuanyuan Xu, Nan Liu, Xia Zhang, Qun Yan, Sheng Zheng, Xiaoyan
Xie, Lijuan Li, Liyun Shen, Hongjie Qian, and Hanxiao Sun for
performing the DNA preparation and experiment.
Conceived and designed the experiments: GN WH BC. Performed the
experiments: XZ MX TG LX. Analyzed the data: GN XZ MX WH HW
BC. Contributed reagents/materials/analysis tools: GN DZ GC Xuejun Li
YB YC YX Xiaoying Li WW WH. Wrote the paper: GN BC XZ MX.
1. Stumvoll M , Goldstein BJ , van Haeften TW ( 2005 ) Type 2 diabetes: principles of pathogenesis and therapy . Lancet 365 : 1333 - 1346 .
2. Zimmet P , Alberti KG , Shaw J ( 2001 ) Global and societal implications of the diabetes epidemic . Nature 414 : 782 - 787 .
3. Yang W , Lu J , Weng J , Jia W , Ji L , et al. ( 2010 ) Prevalence of diabetes among men and women in China . N Engl J Med 362 : 1090 - 1101 .
4. O'Rahilly S ( 2009 ) Human genetics illuminates the paths to metabolic disease . Nature 462 : 307 - 314 .
5. Altshuler D , Hirschhorn JN , Klannemark M , Lindgren CM , Vohl MC , et al. ( 2000 ) The common PPARgamma Pro12Ala polymorphism is associated with decreased risk of type 2 diabetes . Nat Genet 26 : 76 - 80 .
6. Gloyn AL , Weedon MN , Owen KR , Turner MJ , Knight BA , et al. ( 2003 ) Largescale association studies of variants in genes encoding the pancreatic beta-cell KATP channel subunits Kir6.2 (KCNJ11) and SUR1 (ABCC8) confirm that the KCNJ11 E23K variant is associated with type 2 diabetes . Diabetes 52 : 568 - 572 .
7. Grant SF , Thorleifsson G , Reynisdottir I , Benediktsson R , Manolescu A , et al. ( 2006 ) Variant of transcription factor 7-like 2 (TCF7L2) gene confers risk of type 2 diabetes . Nat Genet 38 : 320 - 323 .
8. Sladek R , Rocheleau G , Rung J , Dina C , Shen L , et al. ( 2007 ) A genome-wide association study identifies novel risk loci for type 2 diabetes . Nature 445 : 881 - 885 .
9. Zeggini E , Weedon MN , Lindgren CM , Frayling TM , Elliott KS , et al. ( 2007 ) Replication of genome-wide association signals in UK samples reveals risk loci for type 2 diabetes . Science 316 : 1336 - 1341 .
10. Diabetes Genetics Initiative of Broad Institute of Harvard and MIT, Lund University, and Novartis Institutes of BioMedical Research , Saxena R , Voight BF , Lyssenko V , et al. ( 2007 ) Genome-wide association analysis identifies loci for type 2 diabetes and triglyceride levels . Science 316 : 1331 - 1336 .
11. Scott LJ , Mohlke KL , Bonnycastle LL , Willer CJ , Li Y , et al. ( 2007 ) A genomewide association study of type 2 diabetes in Finns detects multiple susceptibility variants . Science 316 : 1341 - 1345 .
12. Steinthorsdottir V , Thorleifsson G , Reynisdottir I , Benediktsson R , Jonsdottir T , et al. ( 2007 ) A variant in CDKAL1 influences insulin response and risk of type 2 diabetes . Nat Genet 39 : 770 - 775 .
13. Zeggini E , Scott LJ , Saxena R , Voight BF , Marchini JL , et al. ( 2008 ) Metaanalysis of genome-wide association data and large-scale replication identifies additional susceptibility loci for type 2 diabetes . Nat Genet 40 : 638 - 645 .
14. Yasuda K , Miyake K , Horikawa Y , Hara K , Osawa H , et al. ( 2008 ) Variants in KCNQ1 are associated with susceptibility to type 2 diabetes mellitus . Nat Genet 40 : 1092 - 1097 .
15. Unoki H , Takahashi A , Kawaguchi T , Hara K , Horikoshi M , et al. ( 2008 ) SNPs in KCNQ1 are associated with susceptibility to type 2 diabetes in East Asian and European populations . Nat Genet 40 : 1098 - 1102 .
16. Rung J , Cauchi S , Albrechtsen A , Shen L , Rocheleau G , et al. ( 2009 ) Genetic variant near IRS1 is associated with type 2 diabetes, insulin resistance and hyperinsulinemia . Nat Genet 41 : 1110 - 1115 .
17. Kong A , Steinthorsdottir V , Masson G , Thorleifsson G , Sulem P , et al. ( 2009 ) Parental origin of sequence variants associated with complex diseases . Nature 462 : 868 - 874 .
18. Dupuis J , Langenberg C , Prokopenko I , Saxena R , Soranzo N , et al. ( 2010 ) New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk . Nat Genet 42 : 105 - 116 .
19. Saxena R , Hivert MF , Langenberg C , Tanaka T , Pankow JS , et al. ( 2010 ) Genetic variation in gastric inhibitory polypeptide receptor (GIPR) impacts the glucose and insulin responses to an oral glucose challenge . Nat Genet 42 : 142 - 148 .
20. Voight BF , Scott LJ , Steinthorsdottir V , Morris AP , Dina C , et al. ( 2010 ) Twelve type 2 diabetes susceptibility loci identified through large-scale association analysis . Nat Genet 42 : 579 - 589 .
21. Xiang J , Li XY , Xu M , Hong J , Huang Y , et al. ( 2008 ) Zinc transporter-8 gene (SLC30A8) is associated with type 2 diabetes in Chinese . J Clin Endocrinol Metab 93 : 4107 - 4112 .
22. Xu M , Bi Y , Xu Y , Yu B , Huang Y , et al. ( 2010 ) Combined effects of 19 common variations on type 2 diabetes in Chinese: results from two communitybased studies . PLoS One 5 : e14022 .
23. Rosenberg NA , Huang L , Jewett EM , Szpiech ZA , Jankovic I , et al. ( 2010 ) Genome-wide association studies in diverse populations . Nat Rev Genet 11 : 356 - 366 .
24. McCarthy MI ( 2008 ) Casting a wider net for diabetes susceptibility genes . Nat Genet 40 : 1039 - 1040 .
25. Tsai FJ , Yang CF , Chen CC , Chuang LM , Lu CH , et al. ( 2010 ) A genome-wide association study identifies susceptibility variants for type 2 diabetes in Han Chinese . PLoS Genet 6 : e1000847 .
26. Shu XO , Long J , Cai Q , Qi L , Xiang YB , et al. ( 2010 ) Identification of new genetic risk variants for type 2 diabetes . PLoS Genet 6 : e1001127 .
27. Yamauchi T , Hara K , Maeda S , Yasuda K , Takahashi A , et al. ( 2010 ) A genome-wide association study in the Japanese population identifies susceptibility loci for type 2 diabetes at UBE2E2 and C2CD4A-C2CD4B . Nat Genet 42 : 864 - 868 .
28. Xu S , Yin X , Li S , Jin W , Lou H , et al. ( 2009 ) Genomic dissection of population substructure of Han Chinese and its implication in association studies . Am J Hum Genet 85 : 762 - 774 .
29. Xu M , Li XY , Wang JG , Wang XJ , Huang Y , Cheng Q , et al. ( 2009 ) Retinolbinding protein 4 is associated with impaired glucose regulation and microalbuminuria in a Chinese population . Diabetologia 52 : 1511 - 9 .
30. Purcell S , Neale B , Todd-Brown K , Thomas L , Ferreira MA , et al. ( 2007 ) PLINK: a tool set for whole-genome association and population-based linkage analyses . Am J Hum Genet 81 : 559 - 575 .
31. Price AL , Patterson NJ , Plenge RM , Weinblatt ME , Shadick NA , et al. ( 2006 ) Principal components analysis corrects for stratification in genome-wide association studies . Nat Genet 38 : 904 - 909 .
32. Skol AD , Scott LJ , Abecasis GR , Boehnke M ( 2006 ) Joint analysis is more efficient than replication-based analysis for two-stage genome-wide association studies . Nat Genet 38 : 209 - 213 .
33. Mantel N , Haenszel W ( 1959 ) Statistical aspects of the analysis of data from retrospective studies of disease . J Natl Cancer Inst 22 : 719 - 748 .