Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 Apr 1.
Published in final edited form as: Circ Cardiovasc Genet. 2016 Feb 5;9(2):154–161. doi: 10.1161/CIRCGENETICS.115.001246

Interaction of Insulin Resistance and Related Genetic Variants with Triglyceride-Associated Genetic Variants

Yann C Klimentidis 1, Amit Arora 1
PMCID: PMC4838530  NIHMSID: NIHMS758238  PMID: 26850992

Abstract

Background

Several studies suggest that some triglyceride (TG)-associated single nucleotide polymorphisms (SNPs) have pleiotropic and opposite effects on glycemic traits. This potentially implicates them in pathways such as de novo lipogenesis, which is presumably up-regulated in the context of insulin resistance. We therefore tested whether the association of TG-associated SNPs with TG levels differs according to one’s level of insulin resistance.

Methods and Results

In three cohort studies (combined n=12,487), we tested the interaction of established TG-associated SNPs (individually and collectively) with several traits related to insulin resistance, on TG levels. We also tested the interaction of TG SNPs with fasting insulin (FI)-associated SNPs, individually and collectively, on TG levels. We find significant interactions of a weighted genetic risk score (GRS) for TG with insulin resistance on TG (pinteraction=2.73 × 10−11 and pinteraction=2.48 × 10−11 for FI and HOMA-IR, respectively). The association of the TG GRS with TG is over 60% stronger among those in the highest tertile of HOMA-IR, compared to those in the lowest tertile. Individual SNPs contributing to this trend include those in/near GCKR, CILP2, and IRS1, while PIGV-NROB2 and LRPAP1 display an opposite trend of interaction. In the pooled dataset, we also identify a SNP-by-SNP interaction involving a TG-associated SNP, rs4722551 near MIR148A, with a FI-associated SNP, rs4865796 in ARL15 (pinteraction=4.1 × 10−5).

Conclusions

Our findings may thus provide genetic evidence for the upregulation of TG levels in insulin resistant individuals, in addition to identifying specific genetic loci and a SNP-by-SNP interaction implicated in this process.

Keywords: insulin resistance, genetics, genetics, association studies, triglycerides, epistasis, de novo lipogenesis

Introduction

Insulin resistance is a major early risk factor for cardiometabolic disease. The various mechanisms linking insulin resistance with lipid levels and type-2 diabetes (T2D) are still unclear 1.

Through meta-analysis of genome-wide association studies (GWAS), over 30 single-nucleotide polymorphisms (SNPs) have been identified as being associated with triglyceride (TG) levels 2,3. These associations appear to be accentuated in the context of obesity 4–7. However, recent findings suggest that alleles associated with increased TG are also associated with a decreased risk of T2D 8,9, potentially implicating de novo lipogenesis (DNL) in the liver or adipose tissue as one responsible mechanism for this seemingly paradoxical association. Since DNL is likely upregulated in the context of insulin resistance 10–14, converting glucose into fatty acids, it is possible that the ‘TG genes’ which are implicated in DNL are upregulated in the context of insulin resistance.

We tested the hypothesis that the association of TG-associated loci with TG levels is accentuated, independently of adiposity, among those who are in a state of insulin resistance, which we operationalize via various phenotypic and genetic measures. In three cohorts of Americans of European descent, we test the interaction of a genetic risk score for TG, and individual TG-associated loci, with proxy phenotypic and genetic measures of insulin resistance, on TG levels.

Methods

Studies

We used data from three population-based cohort studies in which measurements of triglycerides (TG), high density lipoprotein cholesterol (HDL-C), fasting insulin (FI), fasting glucose (FG), and waist-to-hip ratio (WHR) were available: the Atherosclerosis Risk in Communities (ARIC; n=7,872) 15, the Offspring cohort (Exam 5) of the Framingham Heart Study (FHS; n=2,659) 16,17, and the Multi-Ethnic Study of Atherosclerosis (MESA; n=1,956) 18. We included only individuals who self-reported as White/Caucasian. We obtained approval for this study from the University of Arizona Institutional Review Board, and we obtained data from the database of Genotypes and Phenotypes (dbGaP), through accession numbers: phs000007.v23.p8, phs000280.v2.p1, phs000209.v10.p2.

Phenotypic measures

We excluded subjects with prevalent T2D, which is defined as FG levels >125 mg/dl, a report of clinical diagnosis, or taking any T2D-related medication. We also excluded subjects who did not report fasting for at least 8 hours prior to the baseline exam, or were on cholesterol medications. The blood samples were withdrawn for lipid and glycemic analysis at the baseline exam (at Exam 5, in the case of FHS) and were stored, processed and analyzed using standardized lab protocols and procedures, the details of which have been described elsewhere 19–21. Different units were used by the different studies to report/provide values of TG, HDL-C, FI, and FG (see Table 1). The homeostasis model assessment of insulin resistance (HOMA-IR) 22 was calculated separately in each study as the product of FI (mU/l) and FG (mg/dl), divided by a constant, after converting each of these variables as appropriate (see Supplemental Note). Body mass index (BMI) was calculated as kg/m2, and waist-to-hip ratio (WHR) was calculated as the ratio of waist to hip circumference.

Table 1.

Participant characteristics by study, values are presented as mean ± standard deviation, except for % female.

ARIC (n=7,872) FHS (n=2,659) MESA (n=1,956) p-value for differences across studies
Age (yr) 54 (±6) 54(±10) 62 (±10) <2.20E-16
Sex (% female) 54% 54% 54% 0.91
BMI (kg/m2) 26.7 (±4.6) 27.1 (±4.8) 27.3 (±5.0) 8.24E-10
Waist-hip ratio 0.92 (±0.07) 0.89 (±0.09) 0.91 (±0.08) <2.20E-16
Fasting Insulin (mU/L) 9.99 (±7.09) 8.61 (±7.62) 8.60 (±5.00) 0.03
Fasting glucose (mg/dl) 98.6 (±9.0) 94.9 (±9.6) 87.4 (±9.9) 4.21E-06
HOMA-IR 2.42(±1.92) 2.07 (±1.96) 1.90 (±1.27) 1.12E-03
Plasma triglyceride (mg/dl) 129 (±76) 137 (±96) 127 (±73) 0.05
Plasma HDL-cholesterol (mg/dl) 51.6 (±16.8) 51.2 (±15.2) 53.4 (±16.2) 0.20
TG GRS – 32 SNP 154 (±16) 161 (±17) 155 (±16) < 2.20E-16
FI GRS 0.32(±0.04) 0.31 (±0.04) 0.32(±0.04) 5.43E-03
Open in a new tab

Genotypes and genetic risk scores

Details of study-specific genome-wide genotyping can be found elsewhere 23–25. We performed whole-genome imputation on each dataset separately after standard quality-control procedures. We used IMPUTE2 software and all individuals from the 1,000 Genomes data as reference data 26. Imputation in FHS was performed based on the 1,000 Genomes phase 1 interim release, whereas imputation in MESA and ARIC was performed more recently, and thus based on the more recent phase 1 integrated v3 release. We considered sets of 32 and 40 SNPs robustly associated with TG in two meta-analyses of GWAS 2,3. A list of the 40 SNPs can be found in Supplemental Table 1. Genotypes for all SNPs had imputation accuracy scores (‘info’) > 0.5. The TG GRS in ARIC and MESA is based on 31 SNPs, as there was one SNP (rs2247056) which was not included in the more recent 1,000 Genomes release, possibly due to a change in the rs identifier. We will still refer to this GRS as the 32 SNP GRS for the sake of consistency. Two corresponding weighted genetic risk scores (GRS) were built based on the alleles associated with higher levels of TG and the respective effect size, as determined in the respective GWAS meta-analysis study. The FI GRS was based on 17 SNPs associated with FI, as identified by Scott et al. 27. Briefly, the GRSs were calculated for each individual by summing the number of risk alleles, multiplying the number or estimated dosage (in the case of imputed genotypes) of risk alleles by the corresponding effect size.

Statistical analyses

Differences in demographic and phenotypic characteristics across studies were tested using ANOVA or chi-square test. We used multiple linear regression models to test associations and interactions. Within each study, we log-transformed FI, FG, TG, HDL-C and BMI to approximate a normal distribution and fulfill the assumption of normally distributed residuals. For the analyses in which we combined data across all studies, we also standardized (mean of 0, standard deviation of 1) TG, HDL-C, FG, and FI, separately in each study, prior to combining them. We used sex, age, BMI, HDL-C and WHR (measured at the first exam in ARIC and MESA, and at the fifth exam in FHS) as covariates in all of our statistical models (WHR not included as a covariate when it is the outcome variable), as well as a three-level, categorical, ‘study’ variable, modeled as a random effect, to account for variation across the three studies. We also considered the addition of age-squared as a covariate to account for the potential nonlinear association between age and TG. HDL-C and WHR were included as covariates to avoid confounding since some TG-associated SNPs are also associated with these phenotypes. Interactions were tested by including in the model the product of the TG GRS or TG SNP with either FI, HOMA-IR, FG, or WHR, as well as their respective main effects. Statistical significance of interactions was assessed from the parameter estimate and standard error of the interaction term. For the interaction of TG SNPs with glycemic and insulin-resistance phenotypes, we considered a Bonferroni correction for 40 tests performed, resulting in an alpha=1.25 × 10−3. For the interaction of TG SNPs with FI SNPs, we considered a Bonferroni correction for 520 (40 × 13) pairwise SNPs tests, resulting in an alpha=9.61 × 10−5. We considered this many TG SNP-by-FI SNP interaction tests since some of the TG-FI SNP pairs were highly correlated (r2>0.80), as some loci are both TG- and FI-associated. These include the following FI-associated SNPs (which are also TG-associated loci): rs780094 (GCKR), rs2745353 (RSPO3), rs2943645 (IRS1) and rs731839 (PEPD). Finally, we considered three sensitivity analyses. Since FHS is a family-based study, we considered analyses in which we used pedigree information to include an adjustment for relatedness in a linear mixed model implemented in the package coxme in R Statistical Software 28. We also considered analyses in the absence of adjustment for the covariates BMI, WHR, and HDL-C. Finally, we also conducted interaction analyses using both the 32- and 40-SNP TG GRS. All statistical analyses were conducted with R.

Results

Characteristics of the studies and their participants are shown in Table 1. Mean age and BMI are higher in the MESA cohort, while WHR is lowest in FHS. There are differences in the glycemic and lipid traits across studies, potentially due to slightly different measurement methodologies, among other factors. We first tested the association of each TG GRS with TG (each TG GRS was standardized to facilitate comparison). In the pooled/combined dataset, we find that the association of the 32-SNP TG GRS with TG was stronger than that of the 40-SNP TG GRS (β=0.24 [0.23 – 0.26], p=1.72 × 10−190 vs. β=0.20 [0.18 – 0.22], p=1.86 × 10−128, for the 32- and 40-SNP GRS, respectively). We therefore proceeded with the 32-SNP GRS, although we do consider all 40 SNPs in the single SNP analyses. The FI GRS was strongly associated with FI in the combined dataset (β=2.33 [1.97–2.67], p=1.34 × 10−38). The correlations among the insulin-resistance phenotypes, BMI, HDL-C, and WHR are shown in Supplemental Table 2. Finally, we find that the TG GRS is not significantly associated with HOMA-IR (β=−0.004 [−0.022, −0.013], p=0.62) or with FI (β=−0.001 [−0.019, −0.016], p=0.89), in a model including age, sex, and study as covariates.

The results of the main effect models of age, sex, BMI, WHR, HDL, TG GRS and each of: FI, FG, and HOMA-IR, with TG as the outcome are shown in Supplemental Table 3. Table 2 presents the results of the interaction of each of four phenotypes with the TG GRS on TG levels in all three studies, and in the combined dataset. In all datasets, we find consistently positive interaction coefficients of all four phenotypes with the TG GRS, independently of WHR (for non-WHR outcome phenotypes), BMI, and HDL-C. Based on the interaction coefficients, the interaction appears to be stronger for HOMA-IR and FI than for FG (see Table 2). The strength of the interaction appears to be smaller in MESA, possibly due to the smaller sample size. Adjustment for relatedness in FHS resulted in very slightly attenuated interactions, perhaps due to a reduced sample size (184 individuals without pedigree information). Including only age and sex as covariates, or including age-squared as an additional covariate produced essentially identical results. Using the 40 SNP TG GRS, instead of the 32 SNP GRS, resulted in somewhat attenuated results (βinteraction=0.027, pinteraction=1.12 × 10−4 for HOMA-IR; and βinteraction=0.026, pinteraction=2.16 × 10−4 for FI). Figure 1 shows the association of the TG GRS with TG in three strata (tertiles) of each respective phenotype. The association of the TG GRS with TG was stronger among those individuals with greater insulin resistance, as assessed by HOMA-IR or FI (see Figure 1). For example, the association of the TG GRS with TG among those in the highest tertile of HOMA-IR (β=0.24, p=5.6 × 10−72) was over 60% stronger than among those in the lowest tertile of HOMA-IR (β=0.14, p=4.8 × 10−36). There was no single SNP that drove this interaction, as none reached statistical significance (with same direction of coefficient of interaction) after correction for 40 tests (see Figure 2 and Supplemental Table 1). However, we observed nominally significant interactions for SNPs in/near GCKR, CILP2, and IRS1, and thus these may be strong contributors to the overall GRS trend of interaction with insulin resistance measures. SNPs which exhibited a trend of interaction in the opposite direction (association of SNP with TG weaker in the context of insulin resistance) include those near PIGV-NR0B2 (pinteraction=6.34 × 10−4) and LRPAP1 (see Figure 2 and Supplemental Table 1). Although the interaction of the PIGV-NR0B2 SNP with FI was statistically significant after correction for multiple testing, the direction of interaction was opposite to the direction observed for the TG GRS with FI. Among the SNPs found to show at least a nominally significant trend of interaction with the TG GRS, both LRPAP1 and PIGV/NR0B2 were identified in the most recent large-scale meta-analysis.

Table 2.

Interaction of the 32-SNP TG GRS with each of four phenotypes on TG levels in each individual study, and in a combined/pooled dataset. Covariates: age, sex, BMI, HDL-C, WHR (for non-WHR outcome phenotypes), and study (in combined analysis).

Fasting glucose
Fasting insulin
HOMA-IR
Waist-to-hip ratio
βinteraction pinteraction βinteraction pinteraction βinteraction pinteraction βinteraction pinteraction
ARIC 0.010 0.27 0.042 1.93E-06 0.041 2.80E-06 0.38 9.75E-4
FHS 0.035 0.07 0.074 2.36E-07 0.075 2.32E-07 0.94 2.29E-09
MESA 0.023 0.24 0.037 0.049 0.039 0.038 0.17 0.462
Combined 0.016 0.03 0.046 2.73E-11 0.046 2.48E-11 0.47 3.87E-08
Open in a new tab

HOMA-IR= Homeostasis model assessment of insulin resistance

Figure 1.

Figure 1

Open in a new tab

Association of TG GRS with TG within different strata (tertiles) of each of 4 phenotypes in the combined dataset. β coefficients and corresponding 95% confidence intervals are shown. Age, sex, BMI, HDL-C, WHR (for non-WHR outcome phenotypes), and study are included as covariates.

Figure 2.

Figure 2

Open in a new tab

Interaction of each TG-associated SNP with each corresponding phenotype on TG levels in the combined dataset. Interactions with negative coefficients indicate that the association of the TG-increasing allele with TG is weaker in the presence of higher values of the respective phenotype.

We did not find a significant interaction of the TG GRS with the FI GRS on TG level (βinteraction= −0.005 and pinteraction=0.59). However, upon running pairwise SNP-by-SNP interactions of the individual TG SNPs with the individual FI SNPs, we found a statistically significant interaction of rs4722551 (MIR148A) with rs4865796 (ARL15), whereby the association of rs4722551 with TG is accentuated among those individuals with the FI-increasing allele (A) at rs4865796. MIR148A was identified in the most recent large-scale meta-analysis of lipid levels. As shown in Figure 3, the same interaction trend was observed in all three studies (ARIC: pinteraction=0.003, FHS: pinteraction=0.02, MESA: pinteraction=0.22), and achieved statistical significance in the combined dataset (pinteraction=4.1 × 10−5). Adjustment for relatedness in the FHS dataset resulted in a slightly higher p-value (p=0.08 within FHS), albeit with a smaller sample size, due to 184 individuals without pedigree information. In a model with only age and sex, the p-value of this SNP-by-SNP interaction is 3.99 × 10−4. This attenuation is principally due to the exclusion of HDL-C as a covariate. With age, sex and HDL-C as covariates, the interaction remains statistically significant (pinteraction=5.06 × 10−5). In the combined dataset, the association of MIR148A with TG is stronger in those with the FI-increasing ARL15 genotype (β=0.043, p=0.049) compared to those with the FI-decreasing ARL15 genotype (β= −0.13, p=0.0077).

Figure 3.

Figure 3

Open in a new tab

Association of SNP rs4722551 (MIR148A) with TG levels within three genotype strata of SNP rs4865796 (ARL15) in each of three studies and in the combined dataset. Error bars represent standard errors around the beta coefficients. Models include age, sex, BMI, WHR, HDL-C, and study, as covariates.

Discussion

Across three studies, we find that the association of TG-associated genes with TG is stronger among individuals who are insulin resistant, according to FI and HOMA-IR, independently of BMI, WHR, and HDL-C. We also identify the individual SNPs that may be strong contributors to this interaction, as well as a FI SNP -by- TG SNP interaction.

The fact that the 32 SNP GRS explains more of the TG phenotypic variation than the 40 SNP GRS could be related to the fact that the 32 SNP GRS is based on a meta-analysis in which the ARIC and FHS cohorts were included. However, these studies were a relatively smaller part of the most recent meta-analysis (which was based on 37 additional studies typed with the Metabochip).

Using a genetic approach, our results are consistent with previous studies showing that TG levels are increased in the context of insulin resistance (e.g. 29). Using a very similar approach, Justesen et al. 7 also found an interaction of a TG GRS with HOMA-IR on TG levels in the larger of two studies that they examined. Although no single SNP appeared to drive the observed GRS interaction in our study, SNPs in/near GCKR, CILP2, and IRS1 stand out as potential drivers. The TG-increasing allele of the GCKR variant was also associated with protection against T2D in previous studies 30,31. These results are consistent with a pattern in which insulin resistance is associated with increased glucose uptake and lipogenesis (i.e. DNL), and are consistent with what is known about GCKR, mutations in which can enhance the synthesis of malonyl CoA, a precursor molecule for TG synthesis in the liver 32. Variants in IRS1 have previously been implicated in the insulin cascade, and found to be associated with several cardiometabolic traits 27,33–35. It may be that the association of IRS1 variants with TG is mediated by their association with insulin resistance traits. Little is known about CILP2 and how it might relate to lipid levels. Finally, our result suggesting that the association of rs964184 near APOA1-5 with TG is accentuated in the context of high WHR is consistent with previous studies 4,5.

Another important result is the identification of genes that may have reduced association with TG in the context of insulin resistance. Little is known about how the SNP between PIGV and NR0B2 is related to TG levels. PIGV instructs the synthesis of a glycolipid molecule called glycosyl phosphosphatidyl inositol (GPI) mannosyltransferase, which is known to anchor various proteins on the surface of the cell 36. NR0B2 (also known as SHP1) regulates diverse biological pathways inside the liver, and mutations in it can alter hepatic cholesterol and TG metabolism 37. LRPAP1 codes for the low-density lipoprotein receptor- related protein associated protein 1, which is known to influence cholesterol homeostasis 38,39.

Although we did not find an interaction of the TG GRS with the FI GRS, we did identify an interaction of a TG SNP with a FI SNP, which was consistent across all three studies. This interaction involves a TG-associated SNP near MIR148A and a FI-associated SNP in ARL15. ARL15 has previously been found to also be associated with adiponectin levels 40,41, a hormone secreted by adipose tissue and potentially involved in the regulation of glycemic and lipid levels. ARL15 has also been associated with HDL-C 2. This potentially reinforces the justification for including HDL-C as a covariate in the interaction analysis, particularly for interactions involving SNPs that are also associated with HDL-C. The TG-associated SNP is located near MIR148A, which encodes miRNA-148A, a non-coding RNA which suppresses target mRNAs. miRNA-148A has previously been found to be upregulated in adipogenesis, but downregulated within obese adipocytes 42,43, has been found to be highly expressed in the liver 44, and has been identified in a large-scale GWAS meta-analysis for BMI among African-Americans 45.

Our study is strengthened by the use of multiple phenotypes and genotypes which are indicative of insulin resistance and the use of three datasets to confirm our findings. A limitation of our study is that we are using imperfect surrogates of insulin resistance. Other limitations include that only individuals of European descent were examined, and that the mean age of participants was over 50 years old. Thus, these results may not generalize to other ethnic/racial and age groups. Furthermore, as we are limited to observational data, we cannot assess whether the identified SNP-by-SNP interaction represents an interaction at the molecular level, instead of simply being the result of the corresponding phenotype interactions. However, if this was simply the result of the phenotypic interaction, we might expect it to instead involve SNPs with the greatest effect sizes on the respective phenotypes. According to the respective large-scale meta-analyses, ARL15 has the 8th largest effect size out of 17 FI-associated SNPs, while MIR148A has the 32nd largest effect size out of 40 TG-associated SNPs. Furthermore, microRNAs are good candidates for gene-gene interactions since one of their known functions is the silencing of mRNA either through cleavage, destabilization of an mRNA strand, or repressing the translation of mRNA to proteins, thus greatly influencing gene expression 46. Further detailed studies are needed to thoroughly investigate these potential mechanisms.

In conclusion, our results suggest a complex and important interplay between glycemic and lipid traits that we are able to gain some insight into using a genetic approach. However, it will be important that the genes which are being identified in genetic association and interaction studies are studied in greater detail, preferably in the context of the tissues where they may be exerting their effects on phenotypes.

Supplementary Material

001246 - PAP
001246 - Supplemental Material
CircGenetics_CIRCCVG-2015-001246.xml

Clinical Perspective.

Insulin resistance (IR) is a central and early feature of metabolic and cardiovascular disease. The molecular pathways that link IR and lipid levels are incompletely characterized at the level of genetic sequence variation. This study builds on previous studies showing a seemingly paradoxical inverse association of triglyceride (TG)-increasing alleles with type-2 diabetes risk, potentially implicating de novo lipogenesis (DNL) as one responsible mechanism. Since DNL is up-regulated in the context of IR, we sought to determine if genetic risk for elevated plasma TG is accentuated in the context of IR (phenotypically and genetically assessed). We examined the interaction of a TG genetic risk score (GRS) with IR on TG levels across three cohorts (total n=12,487). We find that the association of the TG GRS with TG is over 60% stronger among those in the highest tertile of IR, compared to those in the lowest tertile. The main drivers of this pattern appear to include GCKR, CILP2, and IRS1. We also find a statistically significant SNP-by-SNP interaction involving a TG-associated SNP near MIR148A, and a FI-associated SNP in ARL15. Our study sheds light from a genetic perspective on the interplay of lipid and glycemic traits, reveals specific genetic variants that may be major players in this interplay, and identifies a replicated SNP-by-SNP interaction. These findings may pave the way towards identifying novel pathways in cardiometabolic disease and/or better characterizing currently identified pathways and mechanisms. Improved risk prediction, targeted therapeutics and prevention strategies may subsequently be realized.

Acknowledgments

The authors thank the participants and organizers of the ARIC, Framingham, and MESA studies. Framingham Heart Study: The Framingham Heart Study is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with Boston University (Contract No. N01-HC-25195). This manuscript was not prepared in collaboration with investigators of the Framingham Heart Study and does not necessarily reflect the opinions or views of the Framingham Heart Study, Boston University, or NHLBI. Funding for SHARe Affymetrix genotyping was provided by NHLBI Contract N02-HL-64278. SHARe Illumina genotyping was provided under an agreement between Illumina and Boston University. Atherosclerosis Risk in Communities: The Atherosclerosis Risk in Communities Study is carried out as a collaborative study supported by National Heart, Lung, and Blood Institute contracts (HHSN268201100005C, HHSN268201100006C, HHSN268201100007C, HHSN268201100008C, HHSN268201100009C, HHSN268201100010C, HHSN268201100011C, and HHSN268201100012C). Funding for GENEVA was provided by National Human Genome Research Institute grant U01HG004402 (E. Boerwinkle). The authors thank the staff and participants of the ARIC study for their important contributions. Multi-Ethnic Study of Atherosclerosis: MESA and the MESA SHARe project are conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with MESA investigators. Support for MESA is provided by contracts N01-HC-95159, N01-HC-95160, N01-HC-95161, N01-HC-95162, N01-HC-95163, N01-HC-95164, N01-HC-95165, N01-HC-95166, N01-HC-95167, N01-HC-95168, N01-HC-95169 and CTSA UL1-RR-024156. MESA Family is conducted and supported by the National Heart, Lung, and Blood Institute (NHLBI) in collaboration with MESA investigators. Support is provided by grants and contracts R01HL071051, R01HL071205, R01HL071250, R01HL071251, R01HL071258, R01HL071259, UL1-RR-025005, by the National Center for Research Resources, Grant UL1RR033176, and the National Center for Advancing Translational Sciences, Grant UL1TR000124. This manuscript was not prepared in collaboration with MESA investigators and does not necessarily reflect the opinions or views of MESA, or the NHLBI. Funding for SHARe genotyping was provided by NHLBI Contract N02-HL-64278. Genotyping was performed at Affymetrix (Santa Clara, California, USA) and the Broad Institute of Harvard and MIT (Boston, Massachusetts, USA) using the Affymetrix Genome-Wide Human SNP Array 6.0.

Funding Sources: Y.C.K. and A.A. are supported by a National Institutes of Health grant: K01DK095032.

Footnotes

Conflict of Interest Disclosures: None.

References

  • 1.Li N, Fu J, Koonen DP, Kuivenhoven JA, Snieder H, Hofker MH. Are hypertriglyceridemia and low HDL causal factors in the development of insulin resistance? Atherosclerosis. 2014;233:130–138. doi: 10.1016/j.atherosclerosis.2013.12.013. [DOI] [PubMed] [Google Scholar]
  • 2.Teslovich TM, Musunuru K, Smith AV, Edmondson AC, Stylianou IM, Koseki M, et al. Biological, clinical and population relevance of 95 loci for blood lipids. Nature. 2010;466:707–713. doi: 10.1038/nature09270. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Willer CJ, Schmidt EM, Sengupta S, Peloso GM, Gustafsson S, Kanoni S, et al. Discovery and refinement of loci associated with lipid levels. Nat Genet. 2013;45:1274–83. doi: 10.1038/ng.2797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Cole CB, Nikpay M, Lau P, Stewart AFR, Davies RW, Wells GA, et al. Adiposity significantly modifies genetic risk for dyslipidemia. J Lipid Res. 2014;55:2416–2422. doi: 10.1194/jlr.P052522. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Wu Y, Marvelle AF, Li J, Croteau-Chonka DC, Feranil AB, Kuzawa CW, et al. Genetic association with lipids in Filipinos: waist circumference modifies an APOA5 effect on triglyceride levels. J Lipid Res. 2013;54:3198–3205. doi: 10.1194/jlr.P042077. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Zubair N, Mayer-Davis EJ, Mendez Ma, Mohlke KL, North KE, Adair LS. Genetic risk score and adiposity interact to influence triglyceride levels in a cohort of Filipino women. Nutr Diabetes. 2014;4:e118. doi: 10.1038/nutd.2014.16. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Justesen JM, Allin KH, Sandholt CH, Borglykke A, Krarup NT, Grarup N, et al. Interactions of Lipid Genetic Risk Scores with Estimates of Metabolic Health in a Danish Population. Circ Cardiovasc Genet. 2015;8:465–472. doi: 10.1161/CIRCGENETICS.114.000637. [DOI] [PubMed] [Google Scholar]
  • 8.Klimentidis YC, Chougule A, Arora A, Frazier-Wood AC, Hsu C-H. Triglyceride-Increasing Alleles Associated with Protection against Type-2 Diabetes. PLoS Genet. 2015;11:e1005204. doi: 10.1371/journal.pgen.1005204. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Li N, van der Sijde MR, Study LC, Bakker SJL, Dullaart RPF, van der Harst P, et al. Pleiotropic Effects of Lipid Genes on Plasma Glucose, HbA1c, and HOMA-IR Levels. Diabetes. 2014;63:3149–3158. doi: 10.2337/db13-1800. [DOI] [PubMed] [Google Scholar]
  • 10.Petersen KF, Dufour S, Savage DB, Bilz S, Solomon G, Yonemitsu S, et al. The role of skeletal muscle insulin resistance in the pathogenesis of the metabolic syndrome. Proc Natl Acad Sci U S A. 2007;104:12587–12594. doi: 10.1073/pnas.0705408104. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Perry RJ, Samuel VT, Petersen KF, Shulman GI. The role of hepatic lipids in hepatic insulin resistance and type 2 diabetes. Nature. 2014;510:84–91. doi: 10.1038/nature13478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Azzout-Marniche D, Bécard D, Guichard C, Foretz M, Ferré P, Foufelle F. Insulin effects on sterol regulatory-element-binding protein-1c (SREBP-1c) transcriptional activity in rat hepatocytes. Biochem J. 2000;350:389–393. [PMC free article] [PubMed] [Google Scholar]
  • 13.Lewis GF, Carpentier A, Adeli K, Giacca A. Disordered Fat Storage and Mobilization in the Pathogenesis of Insulin Resistance and Type 2 Diabetes. Endocr Rev. 2002;23:201–229. doi: 10.1210/edrv.23.2.0461. [DOI] [PubMed] [Google Scholar]
  • 14.Schwarz JM, Linfoot P, Dare D, Aghajanian K. Hepatic de novo lipogenesis in normoinsulinemic and hyperinsulinemic subjects consuming high-fat, low-carbohydrate and low-fat, high-carbohydrate isoenergetic diets. Am J Clin Nutr. 2003;77:43–50. doi: 10.1093/ajcn/77.1.43. [DOI] [PubMed] [Google Scholar]
  • 15.The Aric Investigators. The Atherosclerosis Risk in Communities (ARIC) Study: design and objectives. The ARIC investigators. Am J Epidemiol. 1989;129:687–702. [PubMed] [Google Scholar]
  • 16.DAWBER TR, MEADORS GF, MOORE FE., Jr Epidemiological approaches to heart disease: the Framingham Study. Am J Public Health Nations Health. 1951;41:279–281. doi: 10.2105/ajph.41.3.279. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Feinleib M, Kannel WB, Garrison RJ, McNamara PM, Castelli WP. The Framingham Offspring Study. Design and preliminary data. Prev Med. 1975;4:518–525. doi: 10.1016/0091-7435(75)90037-7. [DOI] [PubMed] [Google Scholar]
  • 18.Bild DE, Bluemke DA, Burke GL, Detrano R, Diez Roux AV, Folsom AR, et al. Multi-ethnic study of atherosclerosis: objectives and design. Am J Epidemiol. 2002;156:871–881. doi: 10.1093/aje/kwf113. [DOI] [PubMed] [Google Scholar]
  • 19.Sharrett AR, Ballantyne CM, Coady SA, Heiss G, Sorlie PD, Catellier D, et al. Coronary heart disease prediction from lipoprotein cholesterol levels, triglycerides, lipoprotein(a), apolipoproteins A-I and B, and HDL density subfractions: The Atherosclerosis Risk in Communities (ARIC) Study. Circulation. 2001;104:1108–1113. doi: 10.1161/hc3501.095214. [DOI] [PubMed] [Google Scholar]
  • 20.Cromwell WC, Otvos JD, Keyes MJ, Pencina MJ, Sullivan L, Vasan RS, et al. LDL Particle Number and Risk of Future Cardiovascular Disease in the Framingham Offspring Study - Implications for LDL Management. J Clin Lipidol. 2007;1:583–592. doi: 10.1016/j.jacl.2007.10.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Holvoet P, Jenny NS, Schreiner PJ, Tracy RP, Jacobs DR. The relationship between oxidized LDL and other cardiovascular risk factors and subclinical CVD in different ethnic groups: The Multi-Ethnic Study of Atherosclerosis (MESA) Atherosclerosis. 2007;194:245–252. doi: 10.1016/j.atherosclerosis.2006.08.002. [DOI] [PubMed] [Google Scholar]
  • 22.Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28:412–419. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
  • 23.Korn JM, Kuruvilla FG, McCarroll SA, Wysoker A, Nemesh J, Cawley S, et al. Integrated genotype calling and association analysis of SNPs, common copy number polymorphisms and rare CNVs. Nat Genet. 2008;40:1253–1260. doi: 10.1038/ng.237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Bielinski SJ, Pankow JS, Li N, Hsu FC, Adar SD, Jenny NS, et al. ICAM1 and VCAM1 polymorphisms, coronary artery calcium, and circulating levels of soluble ICAM-1: The multi-ethnic study of atherosclerosis (MESA) Atherosclerosis. 2008;201:339–344. doi: 10.1016/j.atherosclerosis.2008.02.031. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Fan JB, Gunderson KL, Bibikova M, Yeakley JM, Chen J, Wickham Garcia E, et al. Illumina Universal Bead Arrays. Methods Enzymol. 2006;410:57–73. doi: 10.1016/S0076-6879(06)10003-8. [DOI] [PubMed] [Google Scholar]
  • 26.Howie B, Marchini J, Stephens M. Genotype Imputation with Thousands of Genomes. G3 (Bethesda) 2011;1:457–470. doi: 10.1534/g3.111.001198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Scott R, Lagou V, Welch RP, Wheeler E, Montasser ME, Luan J, et al. Large-scale association analyses identify new loci influencing glycemic traits and provide insight into the underlying biological pathways. Nat Genet. 2012;44:991–1005. doi: 10.1038/ng.2385. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Team RDC. R: A language and environment for statistical computing. 2011 [Google Scholar]
  • 29.Tobey TA, Greenfield M, Kraemer F, Reaven GM. Relationship between insulin resistance, insulin secretion, very low density lipoprotein kinetics, and plasma triglyceride levels in normotriglyceridemic man. Metabolism. 1981;30:165–171. doi: 10.1016/0026-0495(81)90167-0. [DOI] [PubMed] [Google Scholar]
  • 30.Vaxillaire M, Cavalcanti-Proenca C, Dechaume A, Tichet J, Marre M, Balkau B, et al. The common P446L polymorphism in GCKR inversely modulates fasting glucose and triglyceride levels and reduces type 2 diabetes risk in the DESIR prospective general French population. Diabetes. 2008;57:2253–2257. doi: 10.2337/db07-1807. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Sparsø T, Andersen G, Nielsen T, Burgdorf KS, Gjesing AP, Nielsen AL, et al. The GCKR rs780094 polymorphism is associated with elevated fasting serum triacylglycerol, reduced fasting and OGTT-related insulinaemia, and reduced risk of type 2 diabetes. Diabetologia. 2008;51:70–75. doi: 10.1007/s00125-007-0865-z. [DOI] [PubMed] [Google Scholar]
  • 32.Beer NL, Tribble ND, McCulloch LJ, Roos C, Johnson PRV, Orho-Melander M, et al. The P446L variant in GCKR associated with fasting plasma glucose and triglyceride levels exerts its effect through increased glucokinase activity in liver. Hum Mol Genet. 2009;18:4081–4088. doi: 10.1093/hmg/ddp357. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Samani NJ, Erdmann J, Hall AS, Hengstenberg C, Mangino M, Mayer B, et al. Genomewide association analysis of coronary artery disease. N Engl J Med. 2007;357:443–453. doi: 10.1056/NEJMoa072366. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Kilpeläinen TO, Zillikens MC, Stančákova A, Finucane FM, Ried JS, Langenberg C, et al. Genetic variation near IRS1 associates with reduced adiposity and an impaired metabolic profile. Nat Genet. 2011;43:753–760. doi: 10.1038/ng.866. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Rung J, Cauchi S, Albrechtsen A, Shen L, Rocheleau G, Cavalcanti-Proença C, et al. Genetic variant near IRS1 is associated with type 2 diabetes, insulin resistance and hyperinsulinemia. Nat Genet. 2009;41:1110–1115. doi: 10.1038/ng.443. [DOI] [PubMed] [Google Scholar]
  • 36.Kang JY, Hong Y, Ashida H, Shishioh N, Murakami Y, Morita YS, et al. PIG-V involved in transferring the second mannose in glycosylphosphatidylinositol. J Biol Chem. 2005;280:9489–9497. doi: 10.1074/jbc.M413867200. [DOI] [PubMed] [Google Scholar]
  • 37.Boulias K, Katrakili N, Bamberg K, Underhill P, Greenfield A, Talianidis I. Regulation of hepatic metabolic pathways by the orphan nuclear receptor SHP. EMBO J. 2005;24:2624–2633. doi: 10.1038/sj.emboj.7600728. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Gonzalez P, Alvarez R, Reguero JR, Batalla A, Alvarez V, Cortina A, et al. Variation in the lipoprotein receptor-related protein, alpha2-macroglobulin and lipoprotein receptor-associated protein genes in relation to plasma lipid levels and risk of early myocardial infarction. Coron Artery Dis. 2002;13:251–254. doi: 10.1097/00019501-200208000-00001. [DOI] [PubMed] [Google Scholar]
  • 39.Willnow TE, Armstrong SA, Hammer RE, Herz J. Functional expression of low density lipoprotein receptor-related protein is controlled by receptor-associated protein in vivo. Proc Natl Acad Sci U S A. 1995;92:4537–4541. doi: 10.1073/pnas.92.10.4537. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Brent Richards J, Waterworth D, O’Rahilly S, Hivert MF, Loos RJF, Perry JRB, et al. A genome-wide association study reveals variants in ARL15 that influence adiponectin levels. PLoS Genet. 2009;5:e1000768. doi: 10.1371/journal.pgen.1000768. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Dastani Z, Hivert M-F, Timpson N, Perry JRB, Yuan X, Scott RA, et al. Novel Loci for Adiponectin Levels and Their Influence on Type 2 Diabetes and Metabolic Traits: A Multi-Ethnic Meta-Analysis of 45,891 Individuals. PLoS Genet. 2012;8:e1002607. doi: 10.1371/journal.pgen.1002607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Xie H, Lim B, Lodish HF. MicroRNAs induced during adipogenesis that accelerate fat cell development are downregulated in obesity. Diabetes. 2009;58:1050–1057. doi: 10.2337/db08-1299. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Ortega FJ, Moreno-Navarrete JM, Pardo G, Sabater M, Hummel M, Ferrer A, et al. MiRNA expression profile of human subcutaneous adipose and during adipocyte differentiation. PLoS One. 2010;5:e9022. doi: 10.1371/journal.pone.0009022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Gailhouste L, Gomez-Santos L, Hagiwara K, Hatada I, Kitagawa N, Kawaharada K, et al. miR-148a plays a pivotal role in the liver by promoting the hepatospecific phenotype and suppressing the invasiveness of transformed cells. Hepatology. 2013;58:1153–1165. doi: 10.1002/hep.26422. [DOI] [PubMed] [Google Scholar]
  • 45.Monda KL, Chen GK, Taylor KC, Palmer C, Edwards TL, Lange LA, et al. A meta-analysis identifies new loci associated with body mass index in individuals of African ancestry. Nat Genet. 2013;45:690–696. doi: 10.1038/ng.2608. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Bartel DP. MicroRNAs: Genomics, Biogenesis, Mechanism, and Function. Cell. 2004;116:281–297. doi: 10.1016/s0092-8674(04)00045-5. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

001246 - PAP
NIHMS758238-supplement-001246_-_PAP.pdf (1.7MB, pdf)
001246 - Supplemental Material
NIHMS758238-supplement-001246_-_Supplemental_Material.pdf (80.2KB, pdf)
CircGenetics_CIRCCVG-2015-001246.xml
NIHMS758238-supplement-CircGenetics_CIRCCVG-2015-001246_xml.xml (8.7KB, xml)

RESOURCES