Genetic association of ADIPOQ gene variants with type 2 diabetes, obesity and serum adiponectin levels in south Indian population

Gene xxx (2013) xxx–xxx

GENE-39012; No. of pages: 10; 4C:

Contents lists available at ScienceDirect

Gene

j ourna l homepage: www.e lsev ie r .com/ locate /gene

Genetic association of ADIPOQ gene variants with type 2 diabetes, obesityand serum adiponectin levels in south Indian population

Kandaswamy Ramya a,1, Kuppuswamy Ashok Ayyappa a,2, Saurabh Ghosh b,Viswanathan Mohan a, Venkatesan Radha a,⁎a Madras Diabetes Research Foundation-ICMR Advanced Centre for Genomics of Type 2 Diabetes and Dr. Mohan's Diabetes Specialities Centre, Gopalapuram, Chennai 600086, Indiab Human Genetics Unit, Indian Statistical Institute, Kolkata 700 108, India

Abbreviations: BMI, Body Mass Index; SNP, SinglHWE, Hardy–Weinberg Equilibrium; T2DM, Type 2 DDisequilibrium; ADIPOQ, Adiponectin.⁎ Corresponding author at: Department of Molecul

Research Foundation & Dr. Mohan's Diabetes Specialitiesfor Genomics of Type 2 Diabetes, Kallam Anji Reddy CenBio Tech Park, For Women Society, SIPCOT IT Park, SirTel.: +91 44 2747 0608, +91 44 4396 8888; fax: +91

E-mail addresses: [email protected], drmohans@vs(V. Radha).

1 Present address: SRM Research Institute, SRM Univ603203, India.

2 Present address: Centre for Nanotechnology andUniversity, Tanjore 613402, India.

0378-1119/$ – see front matter © 2013 Elsevier B.V. All rhttp://dx.doi.org/10.1016/j.gene.2013.09.012

Please cite this article as: Ramya, K., et al., Gein south Indian population, Gene (2013), htt

a b s t r a c t
a r t i c l e i n f o
Article history:
Accepted 4 September 2013Available online xxxx
Keywords:Adiponectin geneHypoadiponectinemiaObesitySouth Indians

Objective: To investigate the genetic association of eight variants of the adiponectin gene with type 2 diabetesmellitus (T2DM), obesity and serum adiponectin level in the south Indian population.Methods: The study comprised of 1100 normal glucose tolerant (NGT) and 1100 type 2 diabetic, unrelated sub-jects randomly selected from the Chennai Urban Rural Epidemiology Study (CURES), in southern India. Fastingserum adiponectin levels were measured by radioimmunoassay. The variants were screened by polymerasechain reaction-restriction fragment length polymorphism. Linkage disequilibrium was estimated from the esti-mates of haplotype frequencies.Results: Of the 8 variants, four SNPs namely, +276 G/T (rs1501299), −4522 C/T (rs822393), −11365 C/G
(rs266729), and +712 G/A (rs3774261) were significantly associated with T2DM in our study population. The−3971 A/G (rs822396) and −11391 G/A (rs17300539) SNPs' association with T2DM diabetes was mediatedthrough obesity (where the association with type 2 diabetes was lost after adjusting for BMI). There was an in-dependent association of +276 G/T (rs1501299) and −3971 A/G (rs822396) SNPs with generalized obesity and+349 A/G (rs2241767) with central obesity. Four SNPs, −3971 A/G (rs822396), +276 G/T (rs1501299), −4522C/T (rs822393) and Y111H T/C (rs17366743) were significantly associatedwith hypoadiponectinemia. The haplo-types GCCATGAAT and AGCGTGGGT conferred lower risk of T2DM in this south Indian population.Conclusion: The adiponectin gene variants and haplotype contribute to the genetic risk towards the developmentof type 2 diabetes, obesity and hypoadiponectinemia in the south Indian population.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction

The study of adipocytes has burgeoned as a result of the rising inci-dence of diabetes and obesity worldwide. The role of adipose tissue isof great significancewhere it contributes to the pathogenesis of diabetesas well as obesity by secreting a variety of secretory proteins. Amongthem, adiponectin is the major adipocyte secretory protein most abun-dantly found in human plasmawith potent roles in insulin sensitivity in

e Nucleotide Polymorphism;iabetes Mellitus; LD, Linkage

ar Genetics, Madras DiabetesCentre, ICMR Advanced Centretre, Plot No. 20, Golden Jubileeuseri, Chennai 603 103, India.44 2835 0935.nl.net, [email protected]

ersity, Kattankulathur, Chennai

Advance biomaterials, SASTRA

ights reserved.

netic association of ADIPOQ gep://dx.doi.org/10.1016/j.gene

muscle and liver, regulating energy homeostasis and glucose tolerance(Yamauchi et al., 2001). Adiponectin, a product of the ADIPOQ gene(also known as APM1, ACRP30 or GBP28) spans approximately 16 kbwith three exons on chromosome3q27 and has been linked to a suscep-tibility locus for metabolic syndrome, type 2 diabetes and cardiovascu-lar disease (Francke et al., 2001; Vionnet et al., 2000). Adiponectinlevels have a strong genetic component, with heritability estimatedbetween 30% and 50% (Comuzzie et al., 2001). Several genome wideassociation studies (GWAS) among European and Asian populationsidentified ADIPOQ locus as the major gene for variation in the serumadiponectin levels (Heid et al., 2010; Ling et al., 2009).

A number of population based studies reported an association be-tween single nucleotide polymorphisms (SNPs) inADIPOQ gene and cir-culating levels of adiponectin (Hivert et al., 2008;Menzaghi et al., 2007;Vasseur et al., 2002). Low circulating levels of adiponectin and SNPs ofADIPOQ were reported to be associated with insulin resistance, type 2diabetes mellitus (T2DM) and central obesity (Kadowaki et al., 2006;Peters et al., 2013). There are strong evidences to prove the geneticbackground of diabetes in Asian Indians (Mohan, 2004). Asian Indianshave a unique phenotype characterized by increased abdominal obesityand visceral fat despite low body mass index (BMI), hyperinsulinemia

ne variants with type 2 diabetes, obesity and serum adiponectin levels.2013.09.012

http://dx.doi.org/10.1016/j.gene.2013.09.012

mailto:[email protected]




http://www.sciencedirect.com/science/journal/03781119


2 K. Ramya et al. / Gene xxx (2013) xxx–xxx

(Mohan et al., 1986), insulin resistance (Sharp et al., 1987), and dyslip-idemia (Mohan andDeepa, 2004), features that have been referred to asthe “Asian Indian Phenotype” (Joshi, 2003) which results in increasedsusceptibility to T2DM (Abate and Chandalia, 2003). In particular,south Asians have greater predisposition to abdominal obesity and vis-ceral fat (Banerji et al., 1999; Deurenberg et al., 1998;Misra et al., 2001).Furthermore, at any BMI, Indians were shown to have greater insulinresistance compared to other ethnic groups (Stratton et al., 2000). Arecent study reported that though body mass index (BMI) is the mostcommon measure of obesity, waist circumference has been shown tobe a more accurate measure of the distribution of body fat (Brown,2009; Dagan et al., 2013). Hence genetic studies on ADIPOQ variantsgain importance in its relationship with diabetes, obesity and serumadiponectin levels in south Asians.

With this background, the present study has been performed withan objective to investigate the genetic association of the variants inthe ADIPOQ gene with type 2 diabetes, obesity indexes (BMI and waistcircumference) and serum adiponectin levels in the south Indianpopulation.

2. Research design and methods

2.1. Subjects and study design

Using a case–control approach, a total of 2200 unrelated studysubjects were recruited, comprising 1100 controls (normal glucose tol-erant (NGT)) and 1100 cases (T2DM) (975 men and 1225 women,mean age 43 ± 14 years, mean BMI 24.2 ± 4.6 kg/m2) from theChennai Urban Rural Epidemiological Study (CURES), an ongoing epide-miological study conducted on a representative population (N20 years)of Chennai, the fourth largest city in India. Themethodology of the studyhas been published elsewhere (Deepa et al., 2003) and is brieflyoutlined here. In Phase 1 of CURES, 26,001 individuals were recruitedbased on a systematic random sampling technique. Subjects with self-reported diabetes taking drug treatment for diabetes were classified as“known diabetes subjects.” All known diabetes subjects (n = 1529)were invited to visit the center for detailed studies. In addition, every10th individual of the 26,001 individuals without known diabetes wasinvited to undergo oral glucose tolerance tests using a 75-g oral glucoseload (dissolved in 250 ml of water) (Phase 3 of CURES). Those whowere confirmed by oral glucose tolerance test to have 2-h plasma glu-cose value ≥11.1 mmol/l (200 mg/dl) based on World Health Organi-zation (WHO) consulting group criteria were labeled as “newlydetected diabetes subjects” and those with 2-h plasma glucose valueb7.8 mmol/l (140 mg/dl) as being NGT (Alberti and Zimmet, 1998).

2.2. Phenotype measurements

Anthropometric measurements including weight, height, andwaist were obtained using standardized techniques. The BMI wascalculated as weight (in kg) divided by the square of height (in m).Biochemical analyses were done on a Hitachi-912 Auto Analyzer(Hitachi, Mannheim, Germany) using kits supplied by Roche Diagnos-tics (Mannheim). Fasting plasma glucose (glucose oxidase–peroxidasemethod), serum cholesterol (cholesterol oxidase-phenol-4-amino-antipyrene peroxidasemethod), serum triglycerides (glycerol phospha-tase oxidase-phenol-4-amino-antipyrene peroxidase method), andhigh-density lipoprotein cholesterol (direct method; polyethyleneglycol-pretreated enzymes) were measured. Low-density lipoproteincholesterol was calculated using the Friedewald formula (Friedewaldet al., 1972). Glycated hemoglobin (HbA1c) was estimated by high-performance liquid chromatography using a Variant™ machine (Bio-Rad, Hercules, CA, USA). Serum insulin concentration was estimatedusing an enzyme-linked immunosorbent assay (Dako, Glostrup,Denmark). Total serum adiponectin was measured by radioimmunoas-say (cat. no. HADP-61HK; Linco Research, St. Charles, MO, USA) and the

Please cite this article as: Ramya, K., et al., Genetic association of ADIPOQ gein south Indian population, Gene (2013), http://dx.doi.org/10.1016/j.gene

intra- and inter-assay coefficients of variation were 0.38 and 0.74, re-spectively, and the lower detection limit was 1 ng/ml.

Informed consent was obtained from all study participants, and thestudywas approved by theMadras Diabetes Research Foundation Insti-tutional Ethics Committee.

2.3. Definition of risk factors

In the present study, associations between SNPs in the ADIPOQgene and measures of obesity (BMI and waist circumference) as gener-alized (BMI) and central obesity (waist circumference) were investigat-ed to evaluate the contribution of the SNPs towards the genetic riskfor obesity. Generalized obesity was defined according to theWorld Health Organization Asia Pacific Guidelines for Asians asnon-obese (BMI b 25 kg/m2) and obese (BMI ≥ 25 kg/m2) (HealthCommunications Australia Pty Ltd., 2000) and the central obesity wasdefined according to the same guidelines as low risk (men b 90 cm;women b 80 cm) and high risk (men ≥ 90 cm; women ≥ 80 cm)(Health Communications Australia Pty Ltd., 2000; Mohan et al.,2007).

2.4. Genetic analysis

Genomic DNA was extracted from the whole blood by the phenol-chloroformmethod of DNA extraction (Fritsch et al., 1982). Eight ADIPOQSNPs namely, −11391 G/A (rs17300539), −11365 C/G (rs266729) in thepromoter region,−4522 C/T (rs822393),−3971 A/G (rs822396) in intron1,+276 G/T (rs1501299),+349 A/G (rs2241767),+712 G/A (rs3774261)in intron 2, and Y111H T/C (rs17366743) in exon 3 were selected basedon a careful literature review and were included in the study based ontheir previous significant association with type 2 diabetes, obesity andserum adiponectin levels. The SNPs were identified in the NCBI data-base (http://www.ncbi.nlm.nih.gov/SNP) and were genotyped by poly-merase chain reaction on a GeneAmp® PCR system 9700 thermal cycler(Applied Biosystems, Foster City, CA) followed by restriction enzyme di-gestion (New England Biolabs, Inc., Beverly, MA). Primers, restrictionenzymes and the rs numbers with relative positions of the SNPs areshown in Supplementary Table 1. The resulting products were electro-phoresed on a 3% agarose gel. To ensure that the genotyping was of ad-equate quality, we performed random duplicates in 20% of the samples.The assays were performed by a technician who was masked to thephenotype, and there was 99% concordance in the genotyping. Further-more, a few variants were confirmed by direct sequencing with an ABIPrism® 3500 genetic analyzer (Applied Biosystems, Foster City, CA).

2.5. Statistical analysis

Statistical Package for Social Sciences for Windows version 17.0(SPSS, Chicago, IL) was used for statistical analysis. The effects of thevariants on quantitative and categorical variables were analyzed. Allelefrequencies were estimated by gene counting. Agreement with Hardy–Weinberg expectations (HWE) was tested using a χ2 goodness-of-fittest. Comparison of the means between the two groups was analyzedby Student's t test. The χ2 test was used to compare the proportions ofgenotypes or alleles. Analyses for T2DM and NGT are given for an “addi-tive” model in which homozygotes for the major allele (0), heterozy-gotes (1), and homozygotes for the minor allele (2) were coded. One-way analysis of variance was used to compare groups for continuousvariables. Logistic regression analysis was used to identify the risk ofthe genotype combinations for T2DM and obesity. T2DM or obesitywas taken as the dependent variable, and the genotypes were used asthe independent variable. As subjects with diabetes were older andhad higher BMI, we adjusted for age, sex, and BMI in all the logistic re-gression analyses. It is possible that some of our controls, who are youn-ger, will develop diabetes because the prevalence of diabetes increaseswith age. In order to circumvent this problem, age was adjusted for in


http://www.ncbi.nlm.nih.gov/SNP


3K. Ramya et al. / Gene xxx (2013) xxx–xxx

the logistic analysis. Furthermore, to adjust for the possible confoundingeffect of age, sex, and BMI, thesewere included as covariates in the anal-ysis of T2DM. For analyzing the risk of obesity; age, sex, and T2DMwereincluded as covariates. Adiponectin values were log transformed to ob-tain better approximations of the normal distribution prior to analysis.Linkage disequilibrium (LD) and haplotype frequencies were estimat-ed using Haploview software (www.broad.mit.edu/mpg/haploview)(Barrett et al., 2005). A P value b 0.05was considered statistically signif-icant. Significant P values obtained were corrected for multiple testing(Bonferroni correction). Power was estimated using an online posthoc power computation tool (www.dssresearch.com/toolkit/spcalc/power_p2.asp).

3. Results

3.1. Clinical and biochemical characteristics of the study subjects

The study subjects comprised of NGT (n = 1100) and T2DM (n =1100) subjects. Table 1 represents the baseline clinical characteristicsof the study subjects. Compared to NGT subjects, the T2DM subjectswere older, and clinical parameters such as BMI (T2DM 25.3 ±4.3 kg/m2 vs. NGT 23.4 ± 4.7 kg/m2, P b 0.0001), waist circumference(T2DM 91.0 ± 10.1 cm vs. NGT 83.5 ± 12.1 cm, P b 0.0001),fasting plasma glucose (T2DM 9.0 ± 3.9 vs. NGT 4.7 ± 0.4 mmol/l,P b 0.0001), HbA1c (P b 0.0001), and systolic and diastolic blood pres-sures (P b 0.0001) were significantly higher in cases (T2DM) comparedto controls (NGT).

3.2. Association of adiponectin (ADIPOQ) gene variantswith type 2 diabetes

The location of the adiponectin gene variant genotype in the presentstudy is presented in Fig. 1. We first examined the association of eightvariants (+276 G/T (rs1501299), −4522 C/T (rs822393), −11365 C/G(rs266729), −3971 A/G (rs822396), −11391 G/A (rs17300539), +712G/A (rs3774261), +349 A/G (rs2241767) and Y111H T/C (rs17366743)in the ADIPOQ gene with type 2 diabetes (Table 2). The genotypic fre-quencies of the eight SNPs in both NGT and T2DM subjects were inHWE (P N 0.05) (Supplementary Table 2). Among the eight SNPs inthe ADIPOQ gene, six SNPs (rs1501299 G/T, rs822393 C/T, rs266729 C/G,rs822396 A/G, rs17300539 G/A and rs3774261 G/A) were significantly as-sociated with type 2 diabetes under an additive model after adjustingfor age, sex and BMI.

Table 1Clinical and biochemical characteristics of the study subjects.

Parameters Normal glucosetolerant subjects(n = 1100)

Type 2 diabeticsubjects(n = 1100)

P value

n (males/females) 478/622 497/603 –

Age (years) 41 ± 13 51 ± 11 b0.001BMI (kg/m2) 23.4 ± 4.7 25.3 ± 4.3 b0.0001Waist circumference (cm) 83.5 ± 12.1 91.0 ± 10.1 b0.0001Hip circumference (cm) 94.0 ± 10.0 98.0 ± 10.0 b0.0001Systolic blood pressure (mm Hg) 115 ± 17 129 ± 21 b0.0001Diastolic blood pressure (mm Hg) 74 ± 11 78 ± 11 b0.0001Fasting plasma glucose (mmol/l) 4.7 ± 0.4 9.0 ± 3.9 b0.00012 h post load plasma glucose (mmol/l) 5.5 ± 0.05 14.6 ± 3.6 b0.0001Fasting serum insulin (μIU/ml) 8.3 ± 5.6 11.3 ± 7.2 b0.0001Glycated hemoglobin (%) 5.5 ± 0.5 8.7 ± 2.4 b0.0001Serum cholesterol (mg/dl) 173 ± 35 203 ± 43 b0.0001Serum triglycerides (mg/dl) 110 ± 67 179 ± 125 b0.0001HDL cholesterol (mg/dl) 43 ± 10 42 ± 10 0.027LDL cholesterol (mg/dl) 107 ± 30 125 ± 39 b0.0001

Data presented as mean ± SD, HDL—high density lipoprotein, LDL—low densitylipoprotein.


The minor allele frequencies of rs1501299 G/T (18.2%), rs822393 C/T(30.9%), rs266729 C/G (26.7%), and rs822396 A/G (36.3%) SNPswere sig-nificantly higher among T2DM study subjects compared to the NGTstudy subjects (13.4%, p = 0.0001; 25.5%, p = 0.0002; 22.8%, p =0.004; 30.6%, p = 0.0001) respectively. Logistic regression analysisshowed an increased risk towards the susceptibility of T2DM amongthe study subjects with an odds ratio (OR) of 1.64 (95% confidenceinterval (CI): 1.28–2.09, p b 0.0001); 1.84 (95% CI: 1.47–2.30,p b 0.0001); 1.26 (95% CI: 1.03–1.53, p = 0.02) for rs1501299 G/T,rs822393 C/T, and rs266729 C/G SNPs respectively after adjusting forage, sex and BMI. Whereas in the case of rs822396 A/G, the unadjustedOR for the AG genotype was observed to be 1.39 (95% CI: 1.15–1.67,p = 0.001) and the significance was lost after adjusting for age, sexand BMI (1.09 (95% CI: 0.88–1.35), p = 0.42). Contrastingly, theminor allele frequencies of the SNPs rs17300539 G/A and rs3774261G/A were significantly higher among NGT study subjects with 37.2%and 38.0% when compared to the 32.2% (p = 0.0008) and 33.0%(p = 0.0008) in the T2DM study subjects. Logistic regression analy-sis showed a significant OR for the rs17300539 G/A SNP (0.80, (95%CI: 0.67–0.96), p = 0.02) and the significance was lost afteradjusting for age, sex and BMI (1.08, (0.87–1.33), p = 0.47). In thecase of SNP rs3774261 G/A, the adjusted ORwas observed to be statis-tically significant (GA genotype: 0.80 (95% CI: 0.67–0.97), p = 0.02;AA genotype: 0.65 (95% CI: 0.50–0.86) p = 0.002).

3.3. Association of ADIPOQ gene variants with generalized obesity

The genotype and allele frequencies of the eight SNPs in the ADIPOQgene stratified based on BMI as obese and non-obese is shown inTable 3. Two SNPs, rs1501299G/T and rs822396 A/G showed a significantassociation with generalized obesity using the additivemodel adjustingfor age, sex and type 2 diabetes. The minor allele frequencies forrs1501299 G/T and rs822396 A/G SNPs were significantly higher amongthe obese subjects (17.7% and 35.6%) compared to the non-obese sub-jects (14.0%, p = 0.003 and 31.8%, p = 0.01 respectively). Logisticregression analysis revealed significantly increased risk for rs1501299G/T and rs822396 A/G SNPs towards obesity with an adjusted OR of1.47 (95% CI: 1.17–1.84, p = 0.001) for GT and 1.38 (95% CI: 1.02–1.87, p = 0.03) for GG genotype respectively.

3.4. Association of ADIPOQ gene variants with central obesity(waist circumference) among the study subjects

The analysis was done subdividing the total study subjects basedon risk levels with respect to their waist circumference measures.The individuals with a higher waist circumference (men ≥ 90 cm;women ≥ 80 cm) were grouped under a “High Risk group” and thosewith a lower waist circumference (men b 90 cm; women b 80 cm)were grouped under a “Low Risk group”. Among the eight SNPs in theADIPOQ gene, two SNPS rs17300539 G/A and rs2241767 A/G showed asignificant association with central obesity (Table 4). The frequency ofthe GA genotype of the rs17300539 G/A SNP was significantly higheramong the low risk group (45.1%) compared to the high risk group(40.2%, p = 0.03) with an unadjusted OR of 0.79 (95% CI: 0.64–0.97,p = 0.03). The significance was lost after adjusting for age, sex and di-abetes status (0.99 (95% CI: 0.79–1.25, p = 0.97)),whereas in the caseof the rs2241767 A/G SNP, the minor allele frequency of the “G” allelewas significantly higher among the high risk group (19.6%) when com-pared to the low risk group (15.7%, p = 0.003) with an adjusted OR of1.42 (95% CI: 1.12–1.81, p = 0.004) for the AG genotype.

Associations between waist circumference and disease risk factorshave been shown to be sex dependent, for which the total study subjectswere classified based on gender. Only two SNPs rs1501299 G/T andrs2241767 A/G showed a significant associationwithwaist circumferenceamong female study subjects (Table 5). Theminor allele frequency of the“T” allele of rs1501299 G/T and the “G” allele of the rs2241767 A/G SNP


http://www.broad.mit.edu/mpg/haploview

http://www.dssresearch.com/toolkit/spcalc/power_p2.asp

http://www.dssresearch.com/toolkit/spcalc/power_p2.asp


SNP—single nucleotide polymorphism UTR—untranslated region# published by Vimaleswaran et al. (2008)

UTRSignal peptide

Collagen domainGlobular domain

Intron 1 Intron 2

16 Kb

+10211T/G#

-3971 A/G

-4522 C/T-11391 G/A

5' 3'

-11365 C/G

+276 G/T

+349 A/G

+712 G/AY111H T/C

Exon 1 Exon 2 Exon 3ATG

Fig. 1. Location of the genotyped SNPs in the ADIPOQ gene. SNP—single nucleotide polymorphism. UTR—untranslated region. #Published by Vimaleswaran et al. (2008).


was significantly higher among the high risk group (15.5% and 19.4% re-spectively), compared to the low risk group (10.7%, p = 0.002 and15%, p = 0.004 respectively) with an OR of 1.36 (95% CI: 0.99–1.86,

Table 2Association of ADIPOQ gene polymorphisms with Type 2 diabetes.

SNPs NGT subjects Type 2 diabetessubjects

P va

rs1501299 G/T (+276) (n = 889) (n = 829)GG 669 (75.3%) 555 (66.9%) 0.00GT 202 (22.7%) 246 (29.7%) 0.53TT 18 (2.0%) 28 (3.4%) 0.05MAF (T) 13.4% 18.2% 0.00rs822393 C/T (−4522) (n = 1017) (n = 940)CC 570 (56.0%) 454 (48.3%) 0.00CT 375 (36.9%) 391 (41.6%) 0.19TT 72 (7.1%) 95 (10.1%) 0.00MAF (T) 25.5% 30.9% 0.00rs266729 C/G (−11365) (n = 1053) (n = 1073)CC 621 (59.0%) 572 (53.3%) 0.03CG 383 (36.4%) 430 (40.1%) 0.23GG 49 (4.7%) 71 (6.6%) 0.02MAF (G) 22.8% 26.7% 0.00rs822396 A/G (−3971) (n = 1010) (n = 1002)AA 491(48.6%) 400 (39.9%) 0.00AG 420 (41.6%) 476 (47.5%) 0.48GG 99(9.8%) 126 (12.6%) 0.00MAF (G) 30.6% 36.3% 0.00rs17300539 G/A (−11391) (n = 1083) (n = 1010)GG 436 (40.3%) 472(46.7%) 0.02GA 489 (45.2%) 426 (42.2%) 0.16AA 158 (14.6%) 112 (11.1%) 0.00MAF (A) 37.2% 32.2% 0.00rs3774261 G/A (+712) (n = 1052) (n = 1058)GG 413 (39.3%) 483 (45.7%) 0.02GA 479 (45.5%) 452 (42.7%) 0.15AA 160 (15.2%) 123 (11.6%) 0.00MAF (A) 38.0% 33.0% 0.00rs2241767 A/G (+349) (n = 1004) (n = 1023)AA 699 (69.6%) 713 (69.7%) 0.97AG 271 (27.0%) 276 (27.0%) 0.95GG 34 (3.4%) 34 (3.3%) 0.96MAF (G) 16.9% 16.8% 0.98rs17366743 T/C (Y111H) (n = 759) (n = 848)TT 747 (98.4%) 824 (97.2%) 0.12TC 12 (1.6%) 24 (2.8%) –

CC – – –

MAF (C) 0.8% 1.4% 0.13

OR—odds ratio; MAF—minor allele frequency; NGT—normal glucose tolerance.a Adjusted OR—adjusted odds ratio for age, sex and BMI.


p = 0.05) and 1.39 (95% CI: 1.06–1.83, p = 0.01) after adjustingfor age and diabetes whereas among male study subjects, no signifi-cant association was observed with any of the eight variants.

lue Unadjusted OR, 95% CI, P Adjusted ORa, 95% CI, P

06 (GG vs. GT) Reference Reference(GT vs. TT) 1.46 (1.18–1.82), 0.001 1.64 (1.28–2.09), b0.0001(GG vs. TT) 1.87 (1.02–3.42), 0.04 1.81 (0.93–3.51), 0.0701 –

5 (CC vs. CT) Reference Reference9 (CT vs. TT) 1.30 (1.08–1.58), 0.005 1.84 (1.47–2.30), b0.00013 (CC vs. TT) 1.65 (1.19–2.30), 0.003 2.69 (1.83–3.96), b0.000102 –

(CC vs. CG) Reference Reference(CG vs. GG) 1.21 (1.02–1.45), 0.03 1.26 (1.03–1.53), 0.02(CC vs. GG) 1.57 (1.07–2.30), 0.02 1.37 (0.91–2.07), 0.124 –

06 (AA vs. AG) Reference Reference(AG vs. GG) 1.39 (1.15–1.67), 0.001 1.09 (0.88–1.35), 0.423 (AA vs. GG) 1.56 (1.16–2.09), 0.003 1.19 (0.85–1.67), 0.2901 –

(GG vs. GA) Reference Reference(GA vs. AA) 0.80 (0.67–0.96), 0.02 1.08 (0.87–1.33), 0.473 (GG vs. AA) 0.65(0.49–0.86), 0.003 0.91(0.66–1.25), 0.5708 –

(GG vs. GA) Reference Reference(GA vs. AA) 0.98 (0.78–1.20), 0.84 0.80 (0.67–0.97), 0.022 (GG vs. AA) 0.97 (0.71–1.33), 0.87 0.65 (0.50–0.86), 0.00208 –

(AA vs. AG) Reference Reference(AG vs. GG) 0.99 (0.82–1.21), 0.98 0.67 (0.53–0.85), 0.12(AA vs. GG) 0.98 (0.60–1.59), 0.93 0.80 (0.45–1.42), 0.45

–

(TT vs. TC) Reference Reference1.81 (0.89–3.64), 0.09 1.86 (0.90–3.83), 0.09– –



Table 3Association of ADIPOQ gene polymorphisms with generalized obesity among total study subjects.

SNPs Non-obese subjects(b25 kg/m2)

Obese subjects(≥25 kg/m2)

P value Unadjusted OR, 95% CI, P Adjusted ORa, 95% CI, P

rs1501299 G/T (+276) (n = 895) (n = 818)GG 667 (74.5%) 552 (67.5%) 0.002 (GG vs. GT) Reference ReferenceGT 206 (23.0%) 242 (29.6%) 0.93 (GT vs. TT) 1.41 (1.14–1.76), 0.002 1.47 (1.17–1.84), 0.001TT 22 (2.5%) 24 (2.9%) 0.44 (GG vs. TT) 1.31 (0.73–2.37), 0.35 1.42 (0.77–2.60), 0.25MAF (T) 14.0% 17.7% 0.003 –

rs822396 A/G (−3971) (n = 1155) (n = 862)AA 540 (46.8%) 355 (41.2%) 0.03 (AA vs. AG) Reference ReferenceAG 495 (42.9%) 400 (46.4%) 0.55 (AG vs. GG) 1.22 (1.01–1.48), 0.03 1.22 (1.00–1.48), 0.04GG 120 (10.4%) 107 (12.4%) 0.05 (AA vs. GG) 1.35 (1.01–1.81), 0.04 1.38 (1.02–1.87), 0.03MAF (G) 31.8% 35.6% 0.01 –

rs17300539 G/A (−11391) (n = 1223) (n = 865)GG 519 (42.4%) 385 (44.5%) 0.33 (GG vs. GA) Reference ReferenceGA 546 (44.6%) 368 (42.5%) 0.77 (GA vs. AA) 0.90 (0.75–1.09), 0.31 0.96 (0.79–1.17), 0.74AA 158 (12.9%) 112 (12.9%) 0.80 (GG vs. AA) 0.95 (0.72–1.25), 0.74 1.01 (0.76–1.34), 0.92MAF (A) 35.2% 34.2% 0.51 –

rs266729 C/G (−11365) (n = 1222) (n = 896)CC 670 (54.8%) 521 (57.9%) 0.12 (CC vs. CG) Reference ReferenceCG 485 (39.7%) 326 (36.2%) 0.46 (CG vs. GG) 0.86 (0.72–1.03), 0.11 0.85 (0.71–1.03), 0.10GG 67 (5.5%) 49 (5.5%) 0.99 (CC vs. GG) 1.01 (0.69–1.48), 0.92 0.99 (0.67–1.46), 0.96MAF (G) 25.3% 24.0% 0.34 –

rs3774261 G/A (+712) (n = 1223) (n = 876)GG 519 (42.4%) 371 (42.1%) 0.98 (GG vs. GA) Reference ReferenceGA 543 (44.4%) 389 (44.1%) 0.73 (GA vs. AA) 1.00 (0.83–1.20), 0.98 1.05 (0.86–1.27), 0.59AA 161 (13.2%) 116 (13.2%) 0.72 (GG vs. AA) 1.06 (0.80–1.38), 0.67 1.10 (0.83–1.46), 0.49MAF (A) 33.4% 35.9% 0.75 –

rs822393 C/T (−4522) (n = 1116) (n = 825)CC 601 (53.4%) 418 (50.7%) 0.16 (CC vs. CT) Reference ReferenceCT 425 (37.7%) 340 (41.2%) 0.75 (CT vs. TT) 1.15 (0.95–1.39), 0.14 1.14 (0.94–1.39), 0.17TT 90 (8.1%) 67 (8.1%) 0.76 (CC vs. TT) 0.96 (0.69–1.34), 0.82 0.92 (0.65–1.30), 0.66MAF (T) 27.1% 28.7% 0.28 –

rs2241767 A/G (+349) (n = 1211) (n = 901)AA 849 (70.1%) 635 (70.1%) 0.89 (AA vs. AG) Reference ReferenceAG 324 (26.8%) 238 (26.3%) 0.59 (AG vs. GG) 0.98 (0.80–1.19), 0.85 1.02 (0.82–1.25), 0.84GG 38 (3.1%) 28 (3.1%) 0.62 (AA vs. GG) 1.16(0.72–1.87), 0.54 1.28(0.77–2.12), 0.33MAF (G) 16.5% 16.8% 0.85 –

rs17366743 T/C (Y111H) (n = 953) (n = 654)TT 932 (97.8%) 639 (97.7%) 0.95 (TT vs. TC) Reference ReferenceTC 21 (2.2%) 15 (2.30%) 1.04 (0.53–2.03), 0.90 0.93 (0.46–1.88), 0.85CC – – – –

MAF (C) 1.1% 1.1% 0.95 –

OR—odds ratio; MAF—minor allele frequency.a Adjusted OR—adjusted odds ratio for age, sex and type 2 diabetes.


3.5. Association between SNPs in the ADIPOQ gene and quantitative traitsamong the study subjects

We assessed the relation of the 8 SNPs in the ADIPOQ gene to clinicalparameters adjusting for age and sex. Among the NGT study subjects,theminor allele “TT” of the+276 G/T (rs1501299) SNPwas significantlyassociated with an increased hip circumference (98.2 ± 9.7 cm, p =0.001), elevated serum total cholesterol (181.3 ± 33.1 mg/dl, p =0.009), LDL-C (117.6 ± 26.8 mg/dl, p = 0.002), serum triglyceride(125.8 ± 69.5 mg/dl, p = 0.03) and fasting serum insulin levels

Table 4Association of ADIPOQ gene polymorphisms with central obesity (waist circumference) among

SNPs Low risk High risk P v

rs17300539 G/A (−11391) (n = 1461) (n = 565)GG 616 (42.2%) 268 (47.4%) 0.0GA 659 (45.1%) 227 (40.2%) 0.6AA 186 (12.7%) 70 (12.4%) 0.4MAF (A) 35.3% 32.5% 0.0rs2241767 A/G (+349) (n = 1462) (n = 592)AA 1049 (71.8%) 384 (64.9%) 0.0AG 367 (25.1%) 184 (31.1%) 0.9GG 46 (3.1%) 24 (4.1%) 0.2MAF (G) 15.7% 19.6% 0.0

OR—odds ratio, MAF—minor allele frequency.a OR—adjusted for age, sex and diabetes.


(12.1 ± 15.2 μU/ml, p = 0.04), and lower HDL-C levels (42.7 ±10.6 mg/dl, p = 0.003) compared to other genotypes. Similarly, theminor allele “GG” of the−11365 C/G (rs266729) SNP showed a signifi-cant association with increased hip circumference (95.2 ± 11.3 cm,p = 0.01), elevated fasting plasma glucose (86.9 ± 8.8 mg/dl, p =0.001), glycated hemoglobin (5.8 ± 0.6, p = 0.007) and serum triglyc-eride levels (138 ± 92, p = 0.03) compared to other genotypes inthe study subjects. Moreover, the minor allele “GG” of the −3971 A/G(rs822396) SNP showed a significantly increased waist circumfer-ence (85.5 ± 10.1 cm, p = 0.001), elevated fasting plasma glucose

total study subjects.

alue Unadjusted OR, 95% CI, P Adjusted ORa, 95% CI, P

3 (GG vs GA) Reference Reference3 (GA vs AA) 0.79 (0.64–0.97), 0.03 0.99 (0.79–1.25), 0.970 (GG vs AA) 0.86 (0.63–1.17), 0.36 0.99 (0.70–1.40), 0.989 – –

04 (AA vs AG) Reference Reference8 (AG vs GG) 1.37 (1.10–1.69), 0.004 1.42 (1.12–1.81), 0.0041 (AA vs GG) 1.42 (0.85–2.36), 0.17 1.76 (0.97–3.17), 0.0603 – –



Table 5Association of ADIPOQ gene polymorphisms with central obesity (waist circumference) among female study subjects.

SNPs Low risk High risk P value Unadjusted OR, 95% CI, P Adjusted ORa, 95% CI, P

rs1501299 G/T (+276) (n = 521) (n = 457)GG 416 (79.8%) 327 (71.6%) 0.007 (GG vs GT) Reference ReferenceGT 98 (18.8%) 118 (25.8%) 0.63 (GT vs TT) 1.53 (1.13–2.07), 0.006 1.36 (0.99–1.86), 0.05TT 7 (1.3%) 12 (2.6%) 0.15 (GG vs TT) 2.18 (0.84–5.60), 0.10 2.26 (0.86–5.91), 0.09MAF (T) 10.7% 15.5% 0.002rs2241767 A/G (+349) (n = 698) (n = 509)AA 505 (72.3%) 331 (65.0%) 0.02 (AA vs AG) Reference ReferenceAG 177 (25.4%) 158 (31.0%) 0.43 (AG vs GG) 1.36 (1.05–1.75), 0.01 1.39 (1.06–1.83), 0.01GG 16 (2.3%) 20 (3.9%) 0.08 (AA vs GG) 1.90 (0.97–3.73), 0.06 1.80 (0.89–3.63), 0.10MAF (G) 15.0% 19.4% 0.004

OR—odds ratio, MAF—minor allele frequency.a OR—adjusted for age and diabetes.


(85.4 ± 8.5 mg/dl, p = 0.05) and serum total cholesterol levels(183 ± 34.9, p = 0.0001) (data not shown).

3.6. Association between SNPs in the ADIPOQ gene and serum adiponectinlevels among the study subjects

Table 6 presents the analysis of the association of individual ADIPOQSNPs with serum adiponectin levels for type 2 diabetes and generalizedobesity stratified based on the genotype. Among the eight SNPsscreened, the minor alleles of rs822396 A/G, rs1501299 G/T, rs822393C/T SNPs and the “TC” genotype of the rs17366743 T/C SNP were sig-nificantly associated with hypoadiponectinemia among NGT (6.3 ±4.0 μg/ml, p b 0.0001; 2.0 ± 0.5 μg/ml, p b 0.0001; 5.38 ± 4.9 μg/ml,p b 0.0001 and 8.3 ± 4.5 μg/ml, p = 0.006) and non-obese study sub-jects (7.5 ± 4.8 μg/ml, p b 0.0001; 1.6 ± 0.6 μg/ml, p b 0.0001; 3.6 ±3.5 μg/ml, p b 0.0001 and 8.0 ± 4.7 μg/ml, p = 0.01) respectivelywhen compared to themajor alleles. Similarly, among obese study sub-jects the minor alleles of rs822396 A/G (6.9 ± 4.4 μg/ml, p b 0.0001),rs1501299 G/T (1.7 ± 0.5 μg/ml, p b 0.0001), rs266729 C/G (5.9 ±2.2 μg/ml, p = 0.02) and rs822393 C/T (2.7 ± 2.4 μg/ml, p b 0.0001)

Table 6Association between SNPs in the ADIPOQ gene and serum adiponectin levelsa (μg/ml) among t

SNP NGT subjects T2DM s

Genotype Mean ± SD P valueb Mean ±

−11391 G/A (rs17300539) GG 6.6 ± 3.6 b0.0001 6.7 ±GA 8.2 ± 4.2 6.7 ±AA 11.1 ± 5.1 8.5 ±

−11365 C/G (rs266729) CC 11.1 ± 5.5 b0.0001 8.4 ±CG 8.5 ± 2.7 6.9 ±GG 8.4 ± 5.1 5.4 ±

−3971 A/G (rs822396) AA 10.1 ± 6.8 b0.0001 9.3 ±AG 9.4 ± 3.6 8.3 ±GG 6.3 ± 4.0 5.8 ±

+276 G/T (rs1501299) GG 11.6 ± 4.5 b0.0001 11.3 ±GT 5.3 ± 1.0 5.8 ±TT 2.0 ± 0.5 1.5 ±

+712 G/A (rs3774261) GG 7.5 ± 2.7 b0.0001 6.2 ±GA 9.5 ± 4.9 6.3 ±AA 12.1 ± 5.3 9.5 ±

+349 A/G (rs2241767) AA 6.4 ± 0.4 0.001 5.2 ±AG 9.8 ± 2.8 7.4 ±GG 10.2 ± 6.5 9.1 ±

−4522 C/T (rs822393) CC 10.9 ± 4.8 b0.001 10.8 ±CT 5.78 ± 2.3 5.86 ±TT 5.38 ± 4.9 2.32 ±

Y111H T/C (rs17366743) TT 13.8 ± 13.3 0.006 7.7 ±TC 8.3 ± 4.5 7.0 ±

Data presented as Mean ± SD. NGT—normal glucose tolerance, T2DM—type 2 diabetes mellitua Log transformed adiponectin levels are presented.b P value—adjusted for age, sex and BMI.c P value—adjusted for age, sex and diabetes status.


showed a significant associationwith hypoadiponectinemiawhen com-pared to the major alleles.

Further, the minor alleles of rs17300539 G/A, rs3774261 G/A andrs2241767 A/G were significantly associated with elevated serumadiponectin levels among NGT (12.1 ± 5.3 μg/ml, p b 0.0001; 12.1 ±5.3 μg/ml, p b 0.0001 and 10.2 ± 6.5 μg/ml, p = 0.001) and non-obese study subjects (10.1 ± 5.4 μg/ml, p b 0.0001; 11.0 ± 5.2 μg/mlp b 0.0001 and 9.5 ± 3.4 μg/ml, p = 0.02) respectively, when com-pared to the major alleles. A similar association was observed amongthe obese study subjects for the minor alleles of rs17300539G/A (9.4 ± 5.5 μg/ml, p b 0.0001) rs3774261 G/A (10.2 ± 4.5 μg/ml,p b 0.0001) and rs2241767 A/G (9.4 ± 3.2 μg/ml, p b 0.0001) SNPswhen compared to the major alleles.

3.7. LD estimation between ADIPOQ SNPs and haplotype analysis

3.7.1. Haplotype analysis for ADIPOQ gene SNPs in association with Type 2diabetes and obesity

Haplotype analysis was performed to evaluate the combined effectof 8 SNPs in the present study along with the one SNP which has beenreported earlier namely rs17846866 T/G (+10211) in this south Indian

he study subjects.

ubjects Non-obese subjects(b25 kg/m2)

Obese subjects(≥25 kg/m2)

SD P valueb Mean ± SD P valuec Mean ± SD P valuec

3.1 b0.0001 7.1 ± 3.9 b0.0001 6.2 ± 2.6 b0.00012.1 7.4 ± 3.8 7.2 ± 2.55.5 10.1 ± 5.4 9.4 ± 5.55.0 b0.0001 8.6 ± 5.3 0.10 8.1 ± 4.8 0.023.0 8.9 ± 5.0 8.1 ± 4.22.6 7.2 ± 3.6 5.9 ± 2.24.5 b0.0001 9.8 ± 5.8 b0.0001 9.4 ± 5.0 b0.00015.1 8.1 ± 4.0 7.3 ± 3.32.8 7.5 ± 4.8 6.9 ± 4.44.1 b0.0001 11.8 ± 4.6 b0.0001 11.1 ± 3.9 b0.00011.0 5.7 ± 1.0 5.4 ± 1.00.4 1.6 ± 0.6 1.7 ± 0.52.6 b0.0001 7.1 ± 3.0 b0.0001 6.6 ± 3.1 b0.00013.3 9.3 ± 5.4 7.4 ± 3.14.2 11.0 ± 5.2 10.2 ± 4.50.8 b0.0001 6.0 ± 0.7 0.002 5.7 ± 1.1 b0.00014.6 9.1 ± 6.0 8.0 ± 5.03.8 9.5 ± 3.4 9.4 ± 3.24.1 b0.001 11.2 ± 4.8 b0.0001 10.7 ± 4.2 b0.00012.3 6.0 ± 2.6 5.7 ± 2.10.8 3.6 ± 3.5 2.7 ± 2.45.3 0.65 12.4 ± 12.0 0.01 7.5 ± 4.6 0.894.3 8.0 ± 4.7 7.2 ± 4.2

s.



Table 7Haplotype frequencies of ADIPOQ gene variants in normal glucose-tolerant and Type 2 diabetes mellitus study subjects.

Haplotype block(N1% frequency)

Haplotype frequency NGT subjects frequency(n = 1017)

T2DM subjects frequency(n = 940)

P valuea Permuted P valueb

BlockACCATGAGT 0.132 0.134 0.138 0.720 17.270ACCGTGAGT 0.058 0.057 0.062 0.437 10.488ACCATGAAT 0.055 0.064 0.048 0.026 0.629GCCATGAAT 0.052 0.067 0.039 4.20 × 10−5 0.001AGCATGAGT 0.051 0.052 0.053 0.87 20.904GCCATGAGT 0.044 0.052 0.038 0.03 0.787GCCGTGAGT 0.039 0.038 0.042 0.52 12.576ACCATGGGT 0.033 0.028 0.039 0.03 0.766AGCGTGAGT 0.024 0.025 0.023 0.72 17.393ACTATGAGT 0.023 0.023 0.024 0.72 17.302ACCAGGAGT 0.022 0.018 0.028 0.03 0.778ACCGTGAAT 0.019 0.021 0.019 0.70 16.781ACTATGAAT 0.019 0.020 0.018 0.73 17.510GGCATGAGT 0.017 0.019 0.016 0.54 12.876ACCAGGGGT 0.014 0.010 0.019 0.01 0.336GCCGTGAAT 0.014 0.012 0.015 0.41 9.830GCCATGGGT 0.013 0.015 0.012 0.45 10.769ACTGTTAAT 0.013 0.011 0.015 0.28 6.665GCTATGAGT 0.013 0.017 0.009 0.02 0.422AGCGTGGGT 0.012 0.019 0.006 9.31 × 10−5 0.002GGTATGAGT 0.012 0.016 0.008 0.016 0.377ACTGTTAGT 0.011 0.007 0.016 0.013 0.300AGTATGAGT 0.011 0.011 0.011 0.961 23.057ACCATGGAT 0.010 0.012 0.008 0.201 4.819

NGT—normal glucose-tolerant, T2DM—type 2 diabetes mellitus.a P value after multiple corrections (P value × number of tests/number of haplotypes).b Permuted P values were calculated based on 10,000 permutation replications.


population (Vimaleswaran et al., 2008). The benefit of a haplotype-based analysis is that it captures all of the variation across a region,which may improve the ability to detect an association. The haplotype-based analysiswas performed usingHaploview,wherein nine loci haplo-typeswere constructed, and the haplotypeswith frequency N1% in eithercases (T2DM or obese) or controls (NGT or non-obese) were selected.None of these nine loci were in strong pairwise LD (r2 b 0.6) (Supple-mentary Table 3).

Fig. 2. Linkage disequilibrium between the AD


Table 7 represents the haplotype frequencies in NGT (control) andT2DM (case) study subjects of ADIPOQ gene. The haplotype frequencieswere significantly different between cases and controls of the nine lociin the ADIPOQ gene namely, −11391 G/A (rs17300539), −11365 C/G(rs266729), −4522 C/T (rs822393), −3971 A/G (rs822396), +10211T/G (rs17846866), +276 G/T (rs1501299), +349 A/G (rs2241767), +712G/A (rs3774261) and Y111H T/C (rs17366743). Fig. 2 shows the LD struc-ture of the ADIPOQ gene SNPs with respect to type 2 diabetes. The

IPOQ gene variants with respect to T2DM.




frequency of the GCCATGAAT haplotype was higher in NGT subjects(0.067) compared to T2DM (0.039; P = 0.00004) study subjects. Theproportion of AGCGTGGGT haplotype was higher in NGT subjects(0.019) compared to T2DM (0.006, P = 0.00009) study subjects, whilethe proportion of the ACCAGGAGT haplotype was also higher in T2DM(0.028) compared to NGT (0.018, P = 0.03) study subjects. Similarly,the ACCAGGGGT haplotype was significantly higher in T2DM (0.019),compared to the NGT (0.01, P = 0.01) study subjects.

Twenty-four tests were performed, corresponding to the 24 haplo-types satisfying the selection criterion, and a multiple correction wasdone using the Bonferroni test. The association of the haplotypesGCCATGAAT, AGCGTGGGT remained significant as the P value was lessthan the Bonferroni threshold of 0.05/24,which, is 0.002. The other hap-lotypes ACCAGGAGT and ACCAGGGGT could not retain their associationafter multiple testing.

Permutation testing was performed on all the haplotypes, wherein10,000 permutations were computed using Haploview. It was observedthat the difference in the proportion in GCCATGAAT and AGCGTGGGThaplotypes between the cases and controls was significant. None ofthe haplotype frequencies tested was significantly different betweenthe obese and non-obese groups (data not shown). The association ofthese haplotypes with quantitative traits however did not sustain themultiple testing criterion in the present study.

In order to study the combined effect of theADIPOQ SNPs in the pres-ent study, a genetic score was developed to test the association of theSNPs that were significantly associated with type 2 diabetes mellitus(T2DM). Five significant SNPs namely, −11391 G/A (rs17300539),−11365 C/G (rs266729), −4522 C/T (rs822393), +276 G/T(rs1501299) and +712 G/A (rs3774261) were selected and we carriedout an association analysis based on the total number of risk alleles con-sidering all the five SNPs simultaneously. Thus the risk score could varybetween 0 (no risk allele at any of the SNPs) to 10 (two risk alleles ateach of the SNPs). Using age, sex and BMI as covariates, we performeda logistic regression of the T2DM status on the risk score. The resultsare provided in Table 8. As expected, we found that the scorewas highlysignificant in differentiating between T2DM and NGT individuals (p-value b 5 × 10−12). The enhanced risk of T2DM with an increase inBMI was also validated (p-value b 5 × 10−8).

4. Discussion

The adiponectin (ADIPOQ or APM1) gene has been found to be highlyassociated with T2DM, obesity and serum adiponectin levels amongnumerous study populations (Comuzzie et al., 2001; Hara et al., 2002;Heid et al., 2006; Menzaghi et al., 2002; Vasseur et al., 2002;Vimaleswaran et al., 2008; Zacharova et al., 2005). The present studywas performed to determine the contribution of the genetic risk ofADIPOQ gene variants or SNPs towards the development of T2DM, obe-sity and hypoadiponectinemia in the south Indian population. EightSNPs were selected in the present study owing to their positive associ-ation with diabetes and obesity in various ethnic populations. Amongseveral SNPs in the promoter region of the ADIPOQ gene, −11391 G/A(rs17300539) and −11365 C/G (rs266729) SNPs were shown to have arelationship with hypoadiponectinemia and risk of developing type 2diabetes (Hara et al., 2002; Vasseur et al., 2002; Wassel et al., 2010).

Table 8Association analysis using a genetic risk score based on the five associated SNPs.

Estimate (β) Standarderror

Observed value oftest statistics (z)

P value

(Intercept) 6.841 0.630 10.864 2 × 10−16

Score −0.394 0.057 −6.964 3.3 × 10−12

Age −0.068 0.007 −10.113 2 × 10−16

Sex −0.254 0.161 −1.578 0.115BMI −0.093 0.017 −5.589 2.29 × 10−8


In the present study, two promoter variants,−11365 C/G (rs266729)and −11391 G/A (rs17300539) were significantly associated with in-creased susceptibility towards T2DM among the study population andour finding is consistent with a meta-analysis by Gong et al. (2010) inEuropean Whites. A study by Laumen et al. (2009) stressed on the im-portance of the location of these ADIPOQ gene variants in the promoterregion and showed that the promoter activity is modulated byeither of these variants. Similar to the elderly Japanese female studysubjects (Tanimura et al., 2011), we observed a significant associationof the −11365 C/G (rs266729) SNP with higher BMI measuresand lower adiponectin levels among NGT study subjects. However, the−11391 G/A (rs17300539) was found to be significantly associatedwith elevated adiponectin levels and lower impact on obesity mea-sures similar to that of GOLDN study (Warodomwichit et al., 2009).The −4522 C/T (rs822393) SNP conferred two fold higher risk towardsT2DM in our study population. This SNP was also significantly associat-ed with hypoadiponectinemia among NGT study subjects, similar to astudy in theWhite population (Rasmussen-Torvik et al., 2009). We ob-served a significant association between the−3971 A/G (rs822396) var-iant and T2DMwhich is consistent with a recent study among TunisianArabs (Mtiraoui et al., 2012). The variant was also associated withhypoadiponectinemia which serves as one of the characteristic featuresof the “Asian Indian phenotype” and shows the impact of adiponectin inassociation with T2DM. The novel finding of our study is the associationbetween the−3971 A/G (rs822396) SNP and obesity (measure of BMI)which has not been observed in previous studies (Hara et al., 2002;Pischon et al., 2007). The variant −3971 A/G (rs822396) conferringrisk towards T2DM in the present study might have an impact onthe adiponectin gene function (Joshi, 2003) as these intronic varia-tions may give rise to alternatively spliced mRNAs and affect the stabil-ity or processing of the mRNA, respectively. Qiao et al. (2005) hasshown that the DNA encoding the first intron of the ADIPOQ gene con-tains adipocyte-specific regulatory elements highly sensitive to theadipogenic transcription factor C/EBPα that confers significant activityon the ADIPOQ promoter. This intron was shown to contain a 34 bp en-hancer, which regulates adiponectin gene expression in an adipose tis-sue specific manner. Therefore in the present study, the finding of thepolymorphism −3971 A/G (rs822396) close to the enhancer in the re-gion of the first intron is of significance.

We observed no association between the +349 A/G (rs2241767)variant and T2DMwhich is similar to the African Americans andWhites(Rasmussen-Torvik et al., 2009). The GT genotype of the SNP+276 G/T(rs1501299) in the present study conferred 1.64 times higher risk to-wards T2DM after adjusting for age, sex and BMI. However an earlierstudy among Japanese subjects (Hara et al., 2002) showed that theG/G genotype at position 276 was at increased risk for type 2 diabetescompared to T/T genotype. In the present study, the TT genotype of+276G/T (rs1501299) variantwas observed to be significantly associat-ed with elevated fasting serum insulin levels and hypoadiponectinemiaamong the NGT study subjects. While studying the association of the+276 G/T (rs1501299) variant with obesity, the GT genotype was setto confer 1.47 times higher risk towards the development of generalizedobesity. But this association was not replicative enough when NGT sub-jects were studied, which clearly shows the mediation of T2DM for itsrisk towards obesity. The AA genotype of the +712 G/A (rs3774261)SNP conferred 0.65 times lower risk towards the development of type2 diabetes. Very few studies have observed a significant associationwith this variant with respect to T2DM and obesity, however, amongthe Korean population, Cheong et al. (2011) observed a significant asso-ciation with ischemic stroke.

We observed no significant association between the Y111H T/C(rs17366743) SNP of the ADIPOQ gene and T2DMor generalized obesity.Howeverwe observed that the TC genotypewas significantly associatedwith hypoadiponectinemia among the NGT study subjects, which issimilar to French Caucasians (Vasseur et al., 2002), Polish subjects(Kretowski et al., 2005) and in elderly Japanese (Tanimura et al.,




2011).We also performed power calculation and the present study had84%, 62%, 52%, 49%, 30%, 21%, 13% and 11% power to detect the associa-tion of rs822396 A/G, rs822393 C/T, rs3774261 G/A, rs17300539 G/A,rs266729 C/G, rs1501299G/T, rs17846866 T/G and rs2241767 A/G variantswith Type 2 diabetes respectively.

We performed a comprehensive analysis involving all the variantsincluding the previously reported SNP rs17846866 T/G (+10211)(Vimaleswaran et al., 2008), sharing their effects together on two im-portant disease phenotypes, type 2 diabetes and obesity by means ofhaplotype analysis. None of the nine loci tested was in very strongpairwise LD (r2 b 0.6). The haplotype analysis showed that the ACCAGGAGT and ACCAGGGGT haplotypes confer increased risk of T2DM, butthe association diminished after multiple correction (Bonferroni). Incontrast, the GCCATGAAT and AGCGTGGGT haplotypes were signifi-cantly associated with decreased risk of T2DM, which remained signifi-cant even after Bonferroni correction conferring lower risk towards thedevelopment of T2DM.However, no significant haplotype combinationswere observed with respect to obesity among the study subjects. Theassociation of the haplotypes was significant even after multiple correc-tions (Bonferroni) in the present study.

To circumvent the problem of population stratification, we per-formed a case–control study at six unlinked marker loci believed to beunrelated to the disease under study but known to have allelic diversityamong different populations (Devlin et al., 2001). The allele frequencydifference was not statistically significant at any of the loci, indicatingthat the findings in this study were unlikely to be an artifact of popula-tion substructuring.

5. Conclusions

To conclude from the present study, among eight SNPs screenedin the ADIPOQ gene, four SNPs namely, +276 G/T (rs1501299), −4522C/T (rs822393), −11365 C/G (rs266729), and +712 G/A (rs3774261)were significantly associated with type 2 diabetes in our study popula-tion. The association of −3971 A/G (rs822396) and −11391 G/A(rs17300539) SNPs with type 2 diabetes wasmediated through obesity.There was an independent association of +276 G/T (rs1501299) and−3971 A/G (rs822396) SNPs with generalized obesity and +349 A/G(rs2241767) with central obesity. AmongNGT subjects, the−11391 G/A(rs17300539), +712 G/A (rs3774261) and +349 A/G (rs2241767) SNPswere significantly associated with elevated adiponectin levels and theother SNPs were significantly associated with lower serum adiponectinlevels. The haplotypes GCCATGAAT and AGCGTGGGTwere significantlyassociated with lower risk of type 2 diabetes. Hence, additional largerreplication studies are required to validate this novel finding.

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.gene.2013.09.012.

Conflict of interest statement

No competing financial interests exist.

Acknowledgments

This work was funded by the DBT project “Search for susceptibilitygenes for type 2 diabetes in Indians” under “Programme support for re-search in diabetes”. Ramya K was supported by the Council of Scientificand Industrial Research (CSIR). The Chennai Wellingdon CorporateFoundation supported the CURES field studies (CURES-110).

References

Abate, N., Chandalia, M., 2003. The impact of ethnicity on type 2 diabetes. J. Diabetes Com-plications 17, 39–58.


Alberti, K.G., Zimmet, P.Z., 1998. Definition, diagnosis and classification of diabetesmellitus and its complications. Part 1: diagnosis and classification of diabetes mellitusprovisional report of a WHO consultation. Diabet. Med. 15, 539–553.

Banerji, M.A., Faridi, N., Atluri, R., Chaiken, R.L., Lebovitz, H.E., 1999. Body composition, vis-ceral fat, leptin, and insulin resistance in Asian Indian men. J. Clin. Endocrinol. Metab.84, 137–144.

Barrett, J.C., Fry, B., Maller, J., Daly, M.J., 2005. Haploview: analysis and visualization of LDand haplotype maps. Bioinformatics 21, 263–265.

Brown, P., 2009. Waist circumference in primary care. Prim. Care Diabetes 3, 259–261.Cheong, M.Y., Bang, O.S., Cha, M.H., Park, Y.K., Kim, S.H., Kim, Y.J., 2011. Association of the

adiponectin gene variations with risk of ischemic stroke in a Korean population,Yonsei. Med. J. 52, 20–25.

Comuzzie, A.G., et al., 2001. The genetic basis of plasma variation in adiponectin, a globalendophenotype for obesity and the metabolic syndrome. J. Clin. Endocrinol. Metab.86, 4321–4325.

Dagan, S.S., Segev, S., Novikov, I., Dankner, R., 2013. Waist circumference vs body massindex in association with cardiorespiratory fitness in healthy men and women: across sectional analysis of 403 subjects. Nutr. J. 12, 12.

Deepa, M., et al., 2003. The Chennai Urban Rural Epidemiology Study CURES)—study designandmethodology (urban component) (CURES-I). J. Assoc. Physicians India 51, 863–870.

Deurenberg, P., Yap, M., van Staveren, W.A., 1998. Body mass index and percent body fat:a meta analysis among different ethnic groups. Int. J. Obes. Relat. Metab. Disord. 22,1164–1171.

Devlin, B., Roeder, K., Wasserman, L., 2001. Genomic control, a new approach to genetic-based association studies. Theor. Popul. Biol. 60, 155–166.

Francke, S., et al., 2001. A genome-wide scan for coronary heart disease suggests in Indo-Mauritians a susceptibility locus on chromosome 16p13 and replicates linkage withthe metabolic syndrome on 3q27. Hum. Mol. Genet. 10, 2751–2765.

Friedewald, W.T., Levy, J.C.R.I., Fredrickson, D.S., 1972. Estimation of the concentration oflow-density lipoprotein cholesterol in plasma, without use of the preparative ultra-centrifuge. Clin. Chem. 18, 499–502.

Fritsch, E.F., Maniatis, T., Sambrook, J., 1982. Molecular cloning: a laboratory manual. ColdSpring Harbor Laboratory 149–151.

Gong, M., Long, J., Liu, Q., Deng, H.C., 2010. Association of the ADIPOQ rs17360539 andrs266729 polymorphismswith type 2 diabetes: ameta-analysis. Mol. Cell. Endocrinol.325, 78–83.

Hara, K., et al., 2002. Genetic variation in the gene encoding adiponectin is associatedwithan increased risk of type 2 diabetes in the Japanese population. Diabetes 51, 536–540.

Health Communications Australia Pty Ltd., Sydney, 2000. pp. 22–29.Heid, I.M., et al., 2006. Genetic architecture of the APM1 gene and its influence on

adiponectin plasma levels and parameters of the metabolic syndrome in 1,727healthy Caucasians. Diabetes 55, 375–384.

Heid, I.M., et al., 2010. Clear detection of ADIPOQ locus as the major gene for plasmaadiponectin: results of genome-wide association analyses including 4659 Europeanindividuals. Atherosclerosis 208, 412–420.

Hivert, M.F., et al., 2008. Common variants in the adiponectin gene (ADIPOQ) associatedwith plasma adiponectin levels, type 2 diabetes, and diabetes-related quantitativetraits: the Framingham Offspring Study. Diabetes 57, 3353–3359.

Joshi, S.R., 2003. Metabolic syndrome—emerging clusters of the Indian phenotype.J. Assoc. Physicians India 51, 445–446.

Kadowaki, T., Yamauchi, T., Kubota, N., Hara, K., Ueki, K., Tobe, K., 2006. Adiponectin andadiponectin receptors in insulin resistance, diabetes, and the metabolic syndrome.J. Clin. Invest. 116, 1784–1792.

Kretowski, A., Gugala, K., Okruszko, A., Wawrusiewicz-Kurylonek, N., Gorska, M., 2005.Single nucleotide polymorphisms in exon 3 of the adiponectin gene in subjectswith type 2 diabetes mellitus. Rocz. Akad. Med. Bialymst. 50, 148–150.

Laumen, H., et al., 2009. Functional characterization of promoter variants of theadiponectin gene complemented by epidemiological data. Diabetes 58, 984–991.

Ling, H., et al., 2009. Genome-wide linkage and association analyses to identify genesinfluencing adiponectin levels: the GEMS study. Obesity (Silver Spring) 17, 737–744.

Menzaghi, C., et al., 2002. A haplotype at the adiponectin locus is associated with obesityand other features of the insulin resistance syndrome. Diabetes 51, 2306–2312.

Menzaghi, C., Trischitta, V., Doria, A., 2007. Genetic influences of adiponectin on insulinresistance, type 2 diabetes, and cardiovascular disease. Diabetes 56, 1198–1209.

Misra, A., Pandey, R.M., Devi, J.R., Sharma, R., Vikram, N.K., Khanna, N., 2001. High preva-lence of diabetes, obesity and dyslipidaemia in urban slum population in northernIndia. Int. J. Obes. Relat. Metab. Disord. 25, 1722–1729.

Mohan, V., 2004. Why are Indians more prone to diabetes? J. Assoc. Physicians India 52,468–474.

Mohan, V., Deepa, R., 2004. Risk factors for coronary artery disease in Indians. J. Assoc.Physicians India 52, 95–97.

Mohan, V., Sharp, P.S., Cloke, H.R., Burrin, J.M., Schumer, B., Kohner, E.M., 1986. Serum immu-noreactive insulin responses to a glucose load in Asian Indian and European type 2 (non-insulin-dependent) diabetic patients and control subjects. Diabetologia 29, 235–237.

Mohan, V., Deepa, M., Farooq, S., Narayan, K.M., Datta, M., Deepa, R., 2007. Anthropomet-ric cut points for identification of cardiometabolic risk factors in an urban AsianIndian population. Metabolism 56, 961–968.

Mtiraoui, N., Ezzidi, I., Turki, A., Chaieb, A., Mahjoub, T., Almawi, W.Y., 2012. Single-nucleotide polymorphisms andhaplotypes in the adiponectin gene contribute to the ge-netic risk for type 2 diabetes in Tunisian Arabs. Diabetes Res. Clin. Pract. 97, 290–297.

Peters, K.E., et al., 2013. A comprehensive investigation of variants in genes encodingadiponectin (ADIPOQ) and its receptors (ADIPOR1/R2), and their association withserum adiponectin, type 2 diabetes, insulin resistance and the metabolic syndrome.BMC Med. Genet. 14, 15.

Pischon, T., et al., 2007. Single nucleotide polymorphisms at the adiponectin locus and riskof coronary heart disease in men and women. Obesity (Silver Spring) 15, 2051–2060.




http://refhub.elsevier.com/S0378-1119(13)01186-4/rf0085
























































































Qiao, L., et al., 2005. C/EBPalpha regulates human adiponectin gene transcription throughan intronic enhancer. Diabetes 54, 1744–1754.

Rasmussen-Torvik, L.J., Pankow, J.S., Jacobs Jr., D.R., Steinberger, J., Moran, A., Sinaiko, A.R.,2009. The association of SNPs in ADIPOQ, ADIPOR1, and ADIPOR2 with insulin sensi-tivity in a cohort of adolescents and their parents. Hum. Genet. 125, 21–28.

Sharp, P.S., Mohan, V., Levy, J.C., Mather, H.M., Kohner, E.M., 1987. Insulin resistance in pa-tients of Asian Indian and European origin with non-insulin dependent diabetes.Horm. Metab. Res. 19, 84–85.

Stratton, I.M., et al., 2000. Association of glycaemiawithmacrovascular andmicrovascularcomplications of type 2 diabetes (UKPDS 35): prospective observational study. BMJ321, 405–412.

Tanimura, D., et al., 2011. Relation of a common variant of the adiponectin gene to serumadiponectin concentration and metabolic traits in an aged Japanese population. Eur.J. Hum. Genet. 19, 262–269.

Vasseur, F., et al., 2002. Single-nucleotide polymorphism haplotypes in the both proximalpromoter and exon 3 of the APM1 gene modulate adipocyte-secreted adiponectinhormone levels and contribute to the genetic risk for type 2 diabetes in French Cau-casians. Hum. Mol. Genet. 11, 2607–2614.


Vimaleswaran, K.S., et al., 2008. A novel association of a polymorphism in the first intronof adiponectin gene with type 2 diabetes, obesity and hypoadiponectinemia in AsianIndians. Hum. Genet. 123, 599–605.

Vionnet, N., et al., 2000. Genomewide search for type 2 diabetes-susceptibility genes inFrench whites: evidence for a novel susceptibility locus for early-onset diabetes onchromosome 3q27-qter and independent replication of a type 2-diabetes locus onchromosome 1q21-q24. Am. J. Hum. Genet. 67, 1470–1480.

Warodomwichit, D., et al., 2009. ADIPOQ polymorphisms, monounsaturated fatty acids,and obesity risk: the GOLDN study. Obesity (Silver Spring) 17, 510–517.

Wassel, C.L., Pankow, J.S., Jacobs Jr., D.R., Steffes, M.W., Li, N., Schreiner, P.J., 2010. Variantsin the adiponectin gene and serum adiponectin: the Coronary Artery Development inYoung Adults (CARDIA) study. Obesity (Silver Spring) 18, 2333–2338.

Yamauchi, T., et al., 2001. The fat-derived hormone adiponectin reverses insulin resis-tance associated with both lipoatrophy and obesity. Nat. Med. 7, 941–946.

Zacharova, J., Chiasson, J.L., Laakso, M., 2005. The common polymorphisms (single nucle-otide polymorphism [SNP] +45 and SNP +276) of the adiponectin gene predict theconversion from impaired glucose tolerance to type 2 diabetes: the STOP-NIDDMtrial. Diabetes 54, 893–899.








































Documents

Genetic association of ADIPOQ gene variants with type 2 diabetes, obesity and serum adiponectin levels in south Indian population