Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2013 May 1.
Published in final edited form as: Metabolism. 2011 Nov 8;61(5):667–671. doi: 10.1016/j.metabol.2011.09.018

Replication and Meta-analysis of the Gene-Environment Interaction between Body Mass Index and the Interleukin-6 Promoter Polymorphism with Higher Insulin Resistance

Patricia C Underwood 1, Bindu Chamarthi 1, Jonathan S Williams 1, Bei Sun 1, Anand Vaidya 1, Benjamin A Raby 2, Jessica Lasky-Su 2, Paul N Hopkins 3, Gail K Adler 1, Gordon H Williams 1
PMCID: PMC3461261  NIHMSID: NIHMS329750  PMID: 22075267

Abstract

Objective

Insulin resistance (IR) is a complex disorder caused by an interplay of both genetic and environmental factors. Recent studies identified a significant interaction between body mass index (BMI) and the rs1800795 polymorphism of the Interleukin-6 (IL-6) gene that influences both IR and onset of type 2 diabetes mellitus (T2DM) with obese individuals homozygous for the C allele demonstrating the highest level of IR and greatest risk for T2DM. Replication of a gene-environment interaction is important to confirm the validity of the initial finding and extends the generalizability of the results to other populations. Thus, the objective of this study was to replicate this gene-environment interaction on IR in a hypertensive population and perform a meta-analysis with prior published results.

Material and Methods

The replication analysis was performed using Caucasian individuals with hypertension (HTN) from the HyperPATH cohort (N=311), genotyped for rs1800795. Phenotype studies were conducted after participants consumed two diets: high sodium (HS) (200mmol/day) and low sodium (LS) (10mmol/day) for 7 days each. Measurements for plasma glucose, insulin, and IL-6 were obtained after 8 hours of fasting. IR was characterized by the homeostatic model assessment (HOMA-IR).

Results

In HyperPATH, BMI was a significant effect modifier of the relationship between rs1800795 and HOMA-IR; higher BMI was associated with higher HOMA-IR among homozygote CC individuals when compared to major allele G carriers (p=0.003). Further, the meta-analysis in 1028 individuals confirmed the result demonstrating the same significant interaction between rs1800795 and BMI on HOMA-IR (p=1.05×10−6).

Conclusion

This rare replication of a gene-environment interaction extends the generalizability of the results to HTN while highlighting this polymorphism as a marker of IR in obese individuals.

Keywords: Interleukin-6 gene, Hypertension, Obesity, Insulin Resistance

Introduction

Chronic inflammation plays a role in the development of many cardio-metabolic diseases including type 2 diabetes (T2DM), insulin resistance (IR), and hypertension (HTN)[1-3]. The pleiotropic cytokine, interleukin-6 (IL-6), is a major player in the pathophysiology of chronic inflammation [4] and systemic levels of IL-6 are positively associated with T2DM, IR, and HTN [2, 5]. It is evident that the interplay of genetic and environmental factors contributes to the pro-inflammatory process present in cardio-metabolic diseases, however; data on the role of the IL-6 gene in these processes conflicts[6-10]. Recent studies clarify these conflicting results by demonstrating that the association between rs1800795- a guanine (G) to cytosine (C) nucleotide change in the promoter region of the IL-6 gene- and IR is modified by BMI; with the C allele associated with higher IR and T2DM in individuals with obesity [11-13].

Although the co-aggregation of IR and HTN are genetically linked and their association increased in obese individuals [14], the association of rs1800795 with IR has not been demonstrated in HTN. Identifying this relationship in HTN would re-affirm the connection between IL-6, obesity, and HTN in humans and extend the generalizability of this gene-environment interaction on IR. Thus, the objective of this study was to 1) examine the association of rs1800795 with IR in HTN, 2) determine whether BMI modifies this association, and 3) perform a meta-analysis of the HyperPATH data and prior published results [11] to confirm that the rs1800795/BMI interaction on IR exists in a larger population.

Methods and Procedures

HyperPATH Participants

The 311 participants studied were part of the Hypertensive Pathotype (HyperPATH) Protocol. All participants were Caucasian with data available for IL-6 genotype and homeostatic model assessment (HOMA-IR)[15]. Population characteristics are listed in Table 1. Serum IL-6 were available in a sub-set (N=130 high sodium (HS); N=144 low sodium (LS)). Although results from the HyperPATH have been reported previously [14, 16-19] the present analyses are original.

Table 1.

Clinical Characteristics of the HyperPATH Cohort on both High and Low Sodium Diets

HyperPATH Cohort Baseline Characteristics
High Sodium Low Sodium p-value
Age (years) 49.0±8.1 -
Body Mass Index (kg/m2) 28.07 ±3.8 -
Females (percent) 127 (41%) -
Fasting Glucose (mg/dl) 90.0±15 94.0±16 0.006
Fasting Insulin (mg/dl) 7.9±7.2 9.6±8.0 <0.0001
HOMA-IR 1.82±1.73 2.27±2.12 0.0005
Systolic Blood Pressure (mm HG) 146.1±19.3 131.2±17.2 <0.0001
Diastolic Blood Pressure (mm Hg) 86.93±11.14 78.9±9.99 <0.0001
Interleukin-6 (pg/ml) 2.13±1.33 1.9±1.67 0.5
IL-6 G>C (rs1800795) genotype
GG 110(35%) -
GC 142(46%) -
CC 59(19%) -
Open in a new tab

Mean ± SD for normally distributed continuous variables; Median ± Inter-quartile Range for non-normal continuous variables (glucose, insulin, HOMA-IR); percentages for categorical variables.

The protocol was approved by the institutional review boards (IRB) of each site and informed consent was obtained prior to enrollment.

HyperPATH Protocol

Details of this protocol are described previously[14, 17]. In brief, participants completed two diets for 7 days each: High Sodium (HS) (200mmol/day) and Low Sodium (LS) (10mmol/day) with each diet also containing 100mmol/day potassium and 20mmol/day calcium. Anti-hypertensive medications were stopped at least 3 weeks prior to study evaluation. Exclusion criteria included participants with known secondary HTN, T2DM, coronary artery disease, stroke, current tobacco or illicit drug use , or alcohol intake more than 12 ounces per week[16]. On the final day of each diet, participants were admitted to the Human Research Center (HRC). Insulin, glucose, and blood pressure (BP) were measured between 8AM and 10AM as previously described after participants remained fasting and supine overnight[14]. IL-6 levels were also collected concurrently and measured using a quantitative enzyme-linked immunoassay (R&D Systems, Inc., Minneapolis, MN).

Genotyping in the HyperPATH Cohort

DNA was extracted and genotyped as previously described[18]. We analyzed rs1800795 to replicate previous findings related to this variant[11, 12]. Genotyping for this SNP had a completion rate of 97%. Repeat genotyping for 10% of the SNPs on this platform demonstrated concordance with the original genotype call.

Phenotypes Examined

The continuous variable, HOMA-IR, was analyzed as the primary outcome for both the replication analysis and meta-analysis.

Statistical Analyses

Statistical analyses were performed using SAS 9.1 (SAS Institute). A chi-square test evaluated Hardy-Weinberg equilibrium (HWE) for the SNP. HyperPATH population characteristics measured after HS and LS diets were compared using a paired t-test. Non-normally distributed variables (insulin, glucose, HOMA-IR index) are shown as median values with the inter-quartile range and compared using the non-parametric Wilcoxon Ranks test. All linear regression models were performed with the continuous independent variable HOMA-IR as the outcome, and accounted for age, gender, BMI, sibling relatedness, and study site (PROC MIXED). The natural-log of HOMA-IR and IL-6 levels were used to meet normality assumptions. To enable direct comparisons with prior reports, we assumed a dominant genetic model (GG/GC =0, CC =1) [11, 12]. All statistical models conducted as a replication used a 1 sided test. All other statistical tests were 2-sided. Significance is indicated for p<0.05.

Meta-analysis

The meta-analysis was conducted using data collected from 1) HyperPATH and 2) previously published data from the Framingham Heart Study (FHS)[11]. The following information was extracted from both: 1) sample size, 2) p value of the interaction between rs1800795 and BMI on HOMA-IR using a multivariate model. In the FHS, the multivariate model was conducted in males only and accounting for age, BMI, smoking status, physical activity and alcohol use. The multivariate model for HyperPATH is described above.

The meta-analysis was conducted using a weighted z-score method using the METAL software package (http//www.sph.umich.edu/csg/abecasis/metal/). METAL accounts for both the direction of association relative to the reference allele and the sample size of each population. The p values from each study are converted to z scores and a sum of z scores is calculated and weighted by the square root of each study’s sample size. The resulting sum is divided by the square root of the total sample size to obtain an overall z statistic [20].

Results

Genetic Association with HOMA-IR in HyperPATH: Primary Phenotype

In the HyperPATH population, rs1800795 was in HWE and had a minor allele (C allele) frequency of 42 percent. Baseline characteristics demonstrated that sodium intake significantly affected HOMA-IR and BP values, but had not effect on circulating IL-6 levels (Table 1). Multivariate linear regression demonstrated that rs1800795 was independently associated with HOMA-IR on both diets, with higher HOMA-IR observed in the CC genotype (HS p=0.01, LS p=0.01). In a subset of individuals with available serum IL-6 levels, the multivariate model was repeated to include IL-6 levels as a predictor of HOMA-IR. Serum IL-6 levels did not significantly contribute to the variance of HOMA-IR on either diet (p=0.7 HS; p=0.9 LS). Further, when the association of rs1800796 with IL-6 levels was tested, IL-6 levels did not differ significantly by genotype (data not shown).

Genetic Association with HOMA-IR in HyperPATH: Replication of SNP and BMI interaction

To replicate previous findings, we examined whether an interaction between rs1800795 and BMI existed in the HyperPATH cohort. A significant interaction between rs1800795 and BMI existed on both HS (p=0.003) and LS (p=0.004) diets (Figure 1). CC individuals had greater HOMA-IR values than G allele carriers at higher BMI levels (Figure 1). When dichotomized by gender, the HyperPATH population demonstrated similar trends for the SNP/BMI interaction, suggesting gender did not affect the outcome in our population.

Figure 1. The Interaction between rs1800795 and BMI on HOMA-IR in the HyperPATH Cohort on High and Low Sodium Diet.

Figure 1

Open in a new tab

The graph depicts the unadjusted data against the predicted lines for rs1800795 using log HOMA-IR as the dependent variable. The p value depicts the interaction for rs1800795 and BMI accounting for age, gender, body mass index (BMI), sibling relatedness, and study site on A) high sodium and B) low sodium dietary intake.

Since the renin angiotensin aldosterone system (RAAS) is highly implicated in the pathophysiology of HTN and IR and prior investigations demonstrate that LS diet raises HOMA-IR[17], we examined whether dietary sodium influenced the SNP/BMI interaction; however, we observed no difference between diets (Figure 1).

Meta-Analysis

A meta-analysis of the HyperPATH and previously reported FHS cohorts was performed. As expected, a significant interaction between rs1800795 and BMI existed and contributed the variance of HOMA-IR. CC individuals had greater HOMA-IR values than G allele carriers at higher BMI levels (p=1.05×10−6) (Table 2).

Table 2.

Results of the Meta-analysis

Results of Meta-Analysis of rs1800795*BMI Interaction on HOMA-IR
SNP N SNP*BMI interaction p-value Meta p-value
HyperPATH rs1800795 311 0.003a) p=1.05×10−6
FHS rs1800795 717 0.0001B)
Open in a new tab
a)

p-value for interaction of rs1800795 and BMI on HOMA-IR accounting for age, gender, BMI, sibling relatedness, study site and controlling for smoking status, physical activity, and alcohol use.

B)

p- value for interaction of rs1800795 and BMI on HOMA-IR in un-related males accounting for age, BMI, smoking status, physical activity, and alcohol use. FHS= Framingham Heart Study, SNP=single nucleotide polymorphism, BMI=body mass index.

Discussion

Our study demonstrates a significant association between rs1800796 and HOMA-IR in a hypertensive cohort and indicates that this association is modified by BMI. This gene-environment interaction is consistent with findings from prior studies and confirmed via meta-analysis. This study furthers the generalizability of the rs1800796/BMI interaction on the association of IR, and implicates inflammation and obesity as intertwined contributors to IR.

Many studies report a significant association between rs1800796 and IR; however, the allele associated with increased IR differs by population. For example, the G allele has been associated with T2DM in lean males[8] and fasting glucose in a meta-analysis[21] however, a SNP-BMI interaction was never examined. Interestingly, when the study is conducted in a heavier population or when a BMI-SNP interaction is considered, the C allele is associated with IR [11, 12]. We confirm this latter finding in our hypertensive population demonstrating that the BMI-SNP interaction influences IR. Our findings also support previous reports that the C allele is associated with the metabolic syndrome[22, 23]; a population that exhibits both IR and HTN. Further, we demonstrate that similar to other studies, IL-6 levels did not appear to influence this association [11, 12]; albeit, we were likely underpowered to detect this effect as very few circulating IL-6 levels were available in our study.

Due to the associative nature of this study, we are unable to determine the direct mechanism underlying the results. However, it is likely that the known relationship between inflammation, obesity, and IR in humans is involved. Obesity and HTN are associated with a pro-inflammatory state [1, 2] and it has been suggested that increased adiposity may lead to higher IL-6 levels and worse IR [24]. However, this study and others found no relationship between rs1800795 and circulating IL-6 levels [11, 12]. As described previously, it is unknown whether average, peak, or inflammatory levels of IL-6 are contributing to IR and fasting IL-6 measurements may not be appropriate for examining the identified genotype-phenotype relationship[11]. Further, since IL-6 is involved in a multitude of inflammatory processes, it is possible that this SNP is directly related to other genes and/or other inflammatory pathways contributing to IR.

This study has several limitations. First, the sample size is relatively small; however, our findings are consistent with prior reports and the meta-analysis adds power and validity to the finding. The cross-sectional nature of this analysis limits us from drawing conclusions of causality or directionality. Further studies are necessary to determine the functionality of this genetic variant in HTN and whether haplotype analyses demonstrate different results. Strengths of the analysis include 1) use of the HyperPATH cohort representing a distinct hypertensive population with extensive phenotyping including anti-hypertensive medication washout and 2) use of a meta-analysis approach to validate prior findings.

In summary, this study demonstrates that rs1800795 of the IL-6 gene promoter is associated with IR in a Caucasian hypertensive cohort, this association is modified by BMI, and this gene-environment interaction was confirmed via meta-analysis. These findings highlight the roles of adiposity and inflammation in the process of IR. Further, this study implicates this polymorphism as a marker of IR in obese individuals with and without HTN and may potentially identify individuals in whom IR may improve with exercise and/or weight reduction[25]. Using genomic markers to identify and design targeted prevention and treatment strategies for individuals most at risk for IR is essential to implement personalized medicine and hopefully, improve clinical outcomes. The identification of this SNP as a marker for IR in obese individuals from multiple cohorts is the first step towards initiating such targeted care.

Acknowledgements

We thank all other investigators and staff of the HyperPATH Protocol including Nancy Brown Vanderbilt University, Nashville TN, as well as the Clinical Translational Science Center staff, and participants of all protocol sites.

Funding: The project described was supported in part by the following grants: U54LM008748 from the National Library of Medicine, UL1RR025758, Harvard Clinical and Translational Science Center, from the National Center for Research Resources and M01-RR02635, Brigham & Women’s Hospital, General Clinical Research Center, from the National Center for Research Resources. As well as NIH grants HL47651, HL59424, F31 NR011108 (PCU), F32 HL104776-01 (AV), T32HL007609 (BC), K23 HL084236 (JSW), Specialized Center of Research (SCOR) in Molecular Genetics of HTN P50HL055000.

List of Abbreviations

IL-6

Interleukin-6

SNP

Single Nucleotide Polymorphism

HTN

Hypertension

BP

Blood pressure

T2DM

Type 2 diabetes mellitus

RAAS

renin-angiotensin-aldosterone system

IR

insulin resistance

HOMA-IR

homeostatic model assessment

MAF

minor allele frequency

HWE

Hardy-Weinberg expectations

BMI

body mass index

FHS

Framingham Heart Study

G

guanine

C

cytosine

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Author Disclosures: The authors have no conflict of interest to disclose.

Disclosure Statement: The authors have no disclosures.

Author contribution: P.C.U. wrote manuscript, study design and conduct, data collection and analysis, data interpretation. B.C. data collection. J.S.W. study design and conduct, data collection and analysis, data interpretation. B.S.. data collection and analysis. A.V. data collection and data interpretation B.R. study design and conduct, data interpretation. J.L-S. data interpretation. P.N.H. study design and conduct, data collection and analysis G.K.A. data interpretation G.H.W. study design and conduct, data collection and analysis, data interpretation.

References

  • 1.Donath MY, Shoelson SE. Type 2 diabetes as an inflammatory disease. Nat Rev Immunol. 11(2):98–107. doi: 10.1038/nri2925. [DOI] [PubMed] [Google Scholar]
  • 2.Madec S, et al. Pattern of expression of inflammatory markers in adipose tissue of untreated hypertensive patients. J Hypertens. 28(7):1459–65. doi: 10.1097/hjh.0b013e3283388871. [DOI] [PubMed] [Google Scholar]
  • 3.Marsland AL, et al. Systemic inflammation and the metabolic syndrome among middle-aged community volunteers. Metabolism. 59(12):1801–8. doi: 10.1016/j.metabol.2010.05.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Kamimura D, Ishihara K, Hirano T. IL-6 signal transduction and its physiological roles: the signal orchestration model. Rev Physiol Biochem Pharmacol. 2003;149:1–38. doi: 10.1007/s10254-003-0012-2. [DOI] [PubMed] [Google Scholar]
  • 5.Pradhan AD, et al. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. Jama. 2001;286(3):327–34. doi: 10.1001/jama.286.3.327. [DOI] [PubMed] [Google Scholar]
  • 6.Fishman D, et al. The effect of novel polymorphisms in the interleukin-6 (IL-6) gene on IL-6 transcription and plasma IL-6 levels, and an association with systemic-onset juvenile chronic arthritis. J Clin Invest. 1998;102(7):1369–76. doi: 10.1172/JCI2629. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hamid YH, et al. Variations of the interleukin-6 promoter are associated with features of the metabolic syndrome in Caucasian Danes. Diabetologia. 2005;48(2):251–60. doi: 10.1007/s00125-004-1623-0. [DOI] [PubMed] [Google Scholar]
  • 8.Illig T, et al. Significant association of the interleukin-6 gene polymorphisms C-174G and A-598G with type 2 diabetes. J Clin Endocrinol Metab. 2004;89(10):5053–8. doi: 10.1210/jc.2004-0355. [DOI] [PubMed] [Google Scholar]
  • 9.Kubaszek A, et al. The C-174G promoter polymorphism of the IL-6 gene affects energy expenditure and insulin sensitivity. Diabetes. 2003;52(2):558–61. doi: 10.2337/diabetes.52.2.558. [DOI] [PubMed] [Google Scholar]
  • 10.Wypasek E, et al. The increased plasma C-reactive protein and interleukin-6 levels in patients undergoing coronary artery bypass grafting surgery are associated with the interleukin-6-174G > C gene polymorphism. Ann Clin Biochem. 47(Pt 4):343–9. doi: 10.1258/acb.2010.090305. [DOI] [PubMed] [Google Scholar]
  • 11.Herbert A, et al. BMI modifies associations of IL-6 genotypes with insulin resistance: the Framingham Study. Obesity (Silver Spring) 2006;14(8):1454–61. doi: 10.1038/oby.2006.165. [DOI] [PubMed] [Google Scholar]
  • 12.Mohlig M, et al. Body mass index and C-174G interleukin-6 promoter polymorphism interact in predicting type 2 diabetes. J Clin Endocrinol Metab. 2004;89(4):1885–90. doi: 10.1210/jc.2003-031101. [DOI] [PubMed] [Google Scholar]
  • 13.Goyenechea E, Parra D, Martinez JA. Impact of interleukin 6 -174G>C polymorphism on obesity-related metabolic disorders in people with excess in body weight. Metabolism. 2007;56(12):1643–8. doi: 10.1016/j.metabol.2007.07.005. [DOI] [PubMed] [Google Scholar]
  • 14.Raji A, et al. Familial aggregation of insulin resistance and cardiovascular risk factors in hypertension. J Clin Hypertens (Greenwich) 2006;8(11):791–6. doi: 10.1111/j.1524-6175.2006.05761.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Matthews DR, et al. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–9. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
  • 16.Chamarthi B, Williams JS, Williams GH. A mechanism for salt-sensitive hypertension: abnormal dietary sodium-mediated vascular response to angiotensin-II. J Hypertens. 28(5):1020–6. doi: 10.1097/HJH.0b013e3283375974. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Garg R, et al. Low-salt diet increases insulin resistance in healthy subjects. Metabolism. doi: 10.1016/j.metabol.2010.09.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Underwood PC, et al. The relationship between peroxisome proliferator-activated receptor-gamma and renin: a human genetics study. J Clin Endocrinol Metab. 95(9):E75–9. doi: 10.1210/jc.2010-0270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Underwood PC, et al. The association of the angiotensinogen gene with insulin sensitivity in humans: a tagging single nucleotide polymorphism and haplotype approach. Metabolism. doi: 10.1016/j.metabol.2010.12.009. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 26(17):2190–1. doi: 10.1093/bioinformatics/btq340. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Huth C, et al. Joint analysis of individual participants’ data from 17 studies on the association of the IL6 variant -174G>C with circulating glucose levels, interleukin-6 levels, and body mass index. Ann Med. 2009;41(2):128–38. doi: 10.1080/07853890802337037. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Grallert H, et al. IL-6 promoter polymorphisms and quantitative traits related to the metabolic syndrome in KORA S4. Exp Gerontol. 2006;41(8):737–45. doi: 10.1016/j.exger.2006.05.002. [DOI] [PubMed] [Google Scholar]
  • 23.Stephens JW, et al. Association between plasma IL-6, the IL6 -174G>C gene variant and the metabolic syndrome in type 2 diabetes mellitus. Mol Genet Metab. 2007;90(4):422–8. doi: 10.1016/j.ymgme.2006.10.004. [DOI] [PubMed] [Google Scholar]
  • 24.Fernandez-Real JM, et al. Interleukin-6 gene polymorphism and insulin sensitivity. Diabetes. 2000;49(3):517–20. doi: 10.2337/diabetes.49.3.517. [DOI] [PubMed] [Google Scholar]
  • 25.Shih KC, et al. Effects of exercise on insulin sensitivity, inflammatory cytokines, and serum tartrate-resistant acid phosphatase 5a in obese Chinese male adolescents. Metabolism. 59(1):144–51. doi: 10.1016/j.metabol.2009.06.035. [DOI] [PubMed] [Google Scholar]

RESOURCES