Interrelationship of the rs7903146 TCF7L2 gene variant with measures of glucose metabolism and adiposity: the NEO Study
Noordam R1, Zwetsloot CPA1, de Mutsert R2, Mook-Kanamori DO2,3, Lamb HJ4, de Roos A4, de Koning EJP5, Rosendaal FR2,6, Willems van Dijk K6,7,8, van Heemst D1
1) Department of Internal Medicine, section Gerontology and Geriatrics, Leiden University Medical Center, Leiden, the Netherlands
2) Department of Clinical Epidemiology, Leiden University Medical Center, Leiden, the Netherlands
3) Department of Public Health and Primary Care, Leiden University Medical Center, Leiden, the Netherlands
4) Department of Radiology, Leiden University Medical Center, Leiden, the Netherlands 5) Department of Internal Medicine, section Nephrology, Leiden University Medical
Center, Leiden, the Netherlands
6) Einthoven Laboratory for Experimental Vascular Medicine, Leiden University Medical Center, Leiden, the Netherlands
7) Department of Internal Medicine, division of Endocrinology, Leiden University Medical Center, Leiden, the Netherlands
8) Department of Human Genetics, Leiden University Medical Center, Leiden, the Netherlands
Address of correspondence Raymond Noordam PhD
Department of Internal Medicine, section Gerontology and Geriatrics, Leiden University Medical Center,
PO Box 9600, 2300 RC Leiden, the Netherlands Email: r.noordam@lumc.nl
Tel: +31 (0)71 526 5213 Fax: +31 (0)71 526 6912
Running title: “TCF7L2, adiposity, and glucose metabolism”
Number of words manuscript: 3 357 Number of words abstract: 236
Number of tables: 3
Number of figures: 1
Number of references: 45 Number of supplementary tables: 4
Abstract
Background and Aims: We investigated the interrelationship of rs7903146-T in TCF7L2 with measures of glucose metabolism and measures of adiposity.
Methods and Results: This cross-sectional analysis was conducted in 5 744 middle-aged participants (mean (standard deviation [SD]) age is 55.9 (6.0) years) from the Netherlands Epidemiology of Obesity (NEO) Study. Associations between rs7903146-T and Type 2 diabetes mellitus (T2D) were assessed with logistic regression. Additive (per-allele)
associations with measures of glucose metabolism (e.g., fasting insulin) and adiposity (e.g., body mass index [BMI]) were examined with multivariable linear regression. In the total study population, rs7903146-T was associated with a higher risk of T2D (additive odds ratio:
1.42; 95% confidence interval: 1.17; 1.72), and specifically with T2D treated with insulin analogs (2.31 [1.19; 4.46]). After exclusion of participants treated with glucose-lowering medication, rs7903146-T was associated with lower mean insulin concentration (additive mean difference: -0.07 SD [-0.14; 0.00]), but not with higher mean glucose concentration (0.03 SD [-0.01; 0.07]). Furthermore, rs7903146-T was associated with, among other measures of adiposity, a lower mean BMI (-0.04 SD [-0.09; -0.00]), and a lower mean total body fat (-0.04 SD [-0.08; -0.00]). The association between rs7903146-T and T2D increased after adjustment for BMI (odds ratio: 1.51 [1.24; 1.86]); the association between rs7903146-T and fasting insulin diminished after adjustment (-0.05 SD [-0.11; 0.02]).
Conclusion: rs7903146-T is associated with a decreased insulin concentration and increased risk of T2D with opposing effects of adjustment for adiposity.
Key words: TCF7L2, glucose, insulin, diabetes, adiposity
Introduction
The Transcription Factor-7-Like-2 (TCF7L2) gene encodes the transcription factor 4 (TCF4), which is a high mobility group box-containing transcription factor of the Wnt- signaling pathway [1, 2]. While TCF7L2 is highly expressed in multiple tissues, the role of TCF4 in these tissues still needs to be unraveled [3]. Genetic variation in TCF7L2 has been repeatedly identified in Genome-Wide Association Studies (GWAS) on Type 2 Diabetes mellitus (T2D) in populations of both European [4-8] and non-European ancestry [9-11].
Within these studies, the rs7903146-T genetic variant was identified as one of the top hits in relation to an increased risk of T2D [12-14].
The rs7903146-T allele, which is located on one of the TCF7L2 intronic regions, has recently been shown to influence expression of ACSL5, a protein with functions in fatty acid metabolism [15]. Besides the role of rs7903146-T in diabetes risk, studies also found
rs7903146-T to be related to impaired insulin secretion [16-18] and worse therapeutic response to sulfonylurea derivatives [19]. These findings have led to the hypothesis that rs7903146 affects T2D risk through effects on insulin secretion [20]. However, rs7903146-T has also been associated with higher nocturnal glucose levels, which could be interpreted as a reflection of an increased hepatic glucose production or decreased basal insulin production [21], and with decreased hepatic insulin sensitivity [22]. In addition, evidence is not consistent whether TCF7L2 mRNA expression levels in blood, adipose tissue or skeletal muscle are associated with measures of glucose metabolism, including insulin sensitivity [22- 24]. Furthermore, rs7903146-T in TCF7L2 has been associated with a lower body mass index (BMI) [25], although a high BMI is a well-established (casual) risk factor for T2D [26, 27].
Furthermore, the association between rs7903146-T and T2D has only been observed in non- obese individuals [28], and with higher fasting glucose and insulin in non-obese individuals without diabetes mellitus [29].
The results from these studies suggest that TCF7L2 is associated with an increased risk of T2D through β-cell dysfunction independent of adiposity. Nevertheless, it remains unclear whether TCF7L2 has pleiotropic (independent) effects on type 2 diabetes mellitus risk and adiposity, or whether adiposity mediates the association between TCF7L2 and measures of glucose metabolism (e.g., T2D, fasting glucose, fasting insulin, use of insulin medication). Therefore, additional studies are warranted. Within the present study, we aimed to investigate the interrelations between rs7903146-T in TCF7L2 and measures of glucose metabolism and measures of adiposity (e.g., BMI, total body fat, and visceral adipose tissue [VAT]).
Material and methods Study setting
The present study was conducted in the Netherlands Epidemiology of Obesity (NEO) study. The NEO study is a population-based, prospective cohort study including 6 671 individuals aged 45–65 years, with an oversampling of individuals with overweight or obesity. The study design and population are described in more detail elsewhere [30]. In short, men and women living in the greater area of Leiden (in the West of the Netherlands) were invited if they were between 45 and 65 years of age and had a self-reported BMI of 27 kg/m2 or higher. In addition, all inhabitants aged between 45 and 65 years from one
municipality (Leiderdorp) were invited to participate irrespective of their BMI, allowing a reference distribution of BMI. The Medical Ethical Committee of the Leiden University Medical Center (LUMC) approved the study. All participants provided written informed consent.
Study design
The present study is a cross-sectional analysis of baseline data from all participants in the NEO study that had been genotyped. Participants were invited for a baseline visit at the NEO study center of the LUMC after an overnight fast. Prior to the NEO study visit, participants completed a general questionnaire at home about demography, lifestyle and medical history. Participants were asked to bring all medication they were using in the month preceding the baseline study visit and research nurses recorded names and dosages of all medication. Participants came to the research site in the morning, and underwent an extensive physical examination, including anthropometry and fasted blood sampling. In the present analysis, we excluded 929 participants because of missing genotype data, being a first-degree relative, or being of a non-European ancestry. As the effect of rs7903146 on measures of
glucose-insulin homeostasis could not be properly interpreted in individuals treated with glucose-lowering medication, we excluded these individuals from the analyses on measures of glucose-insulin homeostasis.
Genotyping
The data of the rs7903146 polymorphism has been extracted from the whole-genome data, which was genotyped using the Illumina HumanCoreExome chip. The rs7903146 polymorphism was in Hardy-Weinberg equilibrium in the total study population as well as in the random subpopulation of those who underwent MRI measurements (p-values > 0.05).
Measures of glucose-insulin metabolism
The measures of glucose metabolism comprise the presence of diabetes mellitus, use of oral-glucose lowering agents or insulin analogues, fasting glucose and insulin
concentrations, and Homeostatic Model Assessment for Insulin Resistance (HOMA-IR) and beta-cell function (HOMA-B).
Diabetes mellitus was defined as self-reported diabetes, the use of glucose lowering medication or a fasting glucose ≥ 7.0 mmol/L [31]. Use of oral glucose-lowering agents or insulin medication was defined based on a medication inventory at the moment of the study.
Fasting blood samples were taken after an overnight fast of at least 10 hours. Glucose and insulin concentrations were measured in plasma with the Siemens Immulite 2500
(Siemens Healthcare Diagnostics, Breda) and the Roche Modular P800 Analyzer (Roche Diagnostics, Mannheim, Germany) respectively. The HOMA-IR was calculated as the fasting glucose (in mmol/L) * fasting insulin (in mU/L)/22.5 [32]. The HOMA-B was calculated as fasting insulin (in mU/L) * 20/(fasting glucose (in mmol/L) - 3.5 [33].
Measures of adiposity
Body weight and percent body fat were estimated using the Tanita bio impedance balance (TBF-310, Tanita, International division, UK) without shoes and one kilogram was subtracted to correct for the weight of clothing. Total body fat mass was calculated using the body weight and percentage total body fat. BMI was calculated as weight (in kilograms) divided by the square of height (in meters). Waist circumference (in centimeters) was measured mid-way between the border of the lower costal margin and the iliac crest.
Abdominal subcutaneous tissue (aSAT) and VAT were assessed with MRI in
combination with proton magnetic resonance spectroscopy (1H-MRS) of hepatic triglyceride content in a random subgroup without contraindications (most notably metallic devices, claustrophobia or a body circumference of more than 1.70 m; N = 2 236 for MRI
subpopulation; N = 1 822 for 1H-MRS subpopulation). Imaging was performed on a 1.5 Tesla MR system (Philips Medical Systems, Best, the Netherlands) using a turbo spin echo imaging protocol. VAT and aSAT were calculated as the average of 3 transverse images, taken at the level of the fifth lumbar vertebra during a breath-hold [34]. 1H-MRS was used to measure hepatic triglyceride content, as described previously [35]. The hepatic triglyceride content was calculated relative to water with the following formula: (signal amplitude of methylene + methyl)/(signal amplitude of water) × 100 [36].
Statistical analyses
In the NEO study, persons with a BMI of 27 kg/m2 or higher are oversampled. To correctly represent characteristics and associations for the general population [37],
adjustments for the oversampling of participants with a BMI ≥ 27 kg/m2 were made. This was done by weighting individuals towards the BMI distribution of participants from the
Leiderdorp municipality [38], whose BMI distribution was similar to the BMI distribution of
the general Dutch population [39]. All results were based on analyses weighted towards the normal BMI distribution. Consequently, the results apply to a population-based study without oversampling of participants with a BMI ≥ 27 kg/m2. However, as a results of the weighting method, it is not possible to present the number of participants in subgroups of the study populations. Instead, we presented the percentages weighted towards the general population with a normal BMI distribution.
Characteristics of the study population were described as mean (standard deviation), median (interquartile range) or percentage, and are presented for the total study population, subpopulation of individuals with MRI or 1H-MRS data, subpopulation of individuals untreated with glucose-lowering treatment, as well as stratified by rs7903146 genotype (notably rs7903146-CC, -CT and -TT).
Using logistic regression analyses, we calculated odds ratios (ORs) and 95%
confidence intervals of the association between rs7903146 and T2D, T2D treated with oral- glucose lowering agents, T2D treated with insulin analogs and untreated T2D. The
associations between rs7903146 and measures of glucose metabolism and adiposity were examined using multivariable linear regression analyses. To increase the comparability of the effect sizes of the different measures dependent on the rs7903146 genotype, the measures of glucose metabolism and adiposity were Z-score normalized for the analyses. Therefore, results can be interpreted as the mean difference in standard deviation between two groups.
All analyses were adjusted for age, sex, and the first 4 principal components (to adjust for possible population stratification). Participants with treated diabetes mellitus at baseline (oral or insulin) were excluded from all linear regression analyses, including the analyses on adiposity to prevent interference of glucose-lowering treatment in these analyses (for the analyses on adiposity measures, results for the total study population are provided in a supplement table). Note that we included participants with untreated diabetes in these
analyses as glucose and insulin levels from these individuals are unaffected by pharmaceutical intervention and therefore do not bias the results. Additive effects of
rs7903146 were examined by including rs7903146 continuously in the model, and reported as mean difference with 95% confidence interval per additional copy of the rs7903146-T allele.
To obtain insights in the interrelations between rs7903146, adiposity and measures of glucose-insulin homeostasis, we conducted the following analyses: 1) Tested whether there is interaction on a multiplicative scale between the rs7903146 polymorphism and measures of adiposity on measures of glucose-insulin homeostasis. 2) Tested the associations between rs7903146 and measures of glucose-insulin homeostasis after additional adjustment for measures of adiposity (e.g., BMI and waist circumference).
All statistical analyses were performed with STATA statistical software, version 12 (StataCorp LP, College Station, TX, USA). A two-sided p-value below 0.05 was considered statistically significant. Additionally, we adjusted the analyses on the measures of glucose metabolism and adiposity separately for multiple testing using the false-discovery rate.
Results
Characteristics of the study population
The characteristics of the study population (Table 1) are presented for the whole study population as well as stratified by rs7903146 genotype. The total group had a mean (SD) age of 55.9 (6.0) years, a mean (SD) BMI of 26.3 (4.4) kg/m2, and consisted for 55.9%
of women and for 5.4% of participants with T2D. Of the total study population, 2.0% was treated with oral glucose-lowering medication, 0.6% was treated with insulin analogues;
2.8% of the study population fulfilled our criteria for having diabetes mellitus, but did not receive medication. The rs7903146 genotype groups (51% rs7903146-CC, 41% rs7903146- CT, and 8% rs7903146-TT) were similar with respect to age and percentage of women.
Overall, the percentage of T2D was higher and measures of adiposity were lower in
rs7903146-CT and -TT carriers. Similar characteristics were observed for the subpopulation from whom we had MRI or 1H-MRS data (Supplementary Table 1) or for the subpopulation who were not treated with glucose-lowering medication (Supplementary Table 2).
Association between rs7903146 and measures of glucose metabolism
Compared with rs7903146-CC carriers (Table 2), rs7903146-CT carriers had a 1.31 (95% confidence interval [CI]: 1.01; 1.71) times higher risk, and rs7903146-TT carriers had a 2.16 (1.43; 3.26) times higher risk to have diabetes mellitus (additive odds ratio: 1.42 [1.17;
1.72). In addition, rs7903146-T was more strongly associated with T2D treated with insulin analogues (additive odds ratio: 2.31 [1.19; 4.46]) than with T2D treated with oral-glucose lowering agents (additive odds ratio: 1.37 [1.00; 1.88]).
Of the different parameters studied in participants who were not using glucose- lowering treatment, the rs7903146-T allele was associated with a lower fasting insulin (additive mean difference: -0.07 SD [-0.14; 0.00]; although not significantly after correction
for multiple testing) concentration as well as with a lower HOMA-IR (additive mean difference: -0.05 SD [-0.12; 0.02]; although not significantly) and a lower HOMA-B (additive mean difference: -0.10 SD [-0.18; -0.03]), but not with fasting glucose concentration (additive mean difference: 0.03 SD [-0.01; 0.07]).
Association between rs7903146 and measures of adiposity
In participants who did not use glucose-lowering treatment, rs7903146-T was associated with lower BMI, waist circumference, total body fat, and body weight (Table 3).
For example, per additional copy of the rs7903146-T allele, body mass index was 0.04 SD (95%CI: -0.09; -0.00; although not significantly after correction for multiple testing) and 0.08 SD (95%CI: -0.19; 0.02; although not significantly) lower respectively for rs7903146-CT and rs7903146-TT carriers, in comparison with rs7903146-CC carriers. Similar associations were observed for waist circumference, total body fat, and body weight. In the random
subpopulation of participants with data from MRI and 1H-MRS, the effect sizes (e.g., the mean difference in VAT/SAT between rs7903146-CC and –TT carriers) were larger than the effect sizes observed with BMI, although statistical evidence for this observation was lacking (p-values for the additive model > 0.05). These results were similar but somewhat weaker after inclusion of participants with T2D (Supplementary Table 3).
Within our study population, there was no interaction between rs7903146-T and measures of adiposity in the analyses on glucose metabolism (p-values for interactions >0.05;
Supplementary Table 4).
After the addition of BMI to the statistical model, the additive OR of rs7903146 on T2D of 1.41 (1.17; 1.72) increased to 1.51 (1.25; 1.86) (Figure 1A). However, the
standardized difference in fasting insulin of -0.07 SD (-0.14; 0.00) attenuated to -0.05 SD (- 0.11; 0.02) after adding BMI to the regression model (Figure 1B). Similar, the standardized
difference in HOMA-B of -0.10 SD (-0.18; -0.03) attenuated to -0.08 SD (-0.16; -0.01) after adding BMI to the regression model (Figure 1C). Results were similar after adding waist circumference or other measures of adiposity to the regression models (results not shown).
Discussion
In this study, we aimed to investigate the interrelationship of rs7903146-T in TCF7L2 with measures of glucose metabolism and measures of adiposity. In participants of the NEO Study, we replicated the association between rs7903146-T and an increased risk of T2D, and showed that rs7903146-T was more strongly associated with T2D treated with insulin
analogs. Carriers of rs7903146-T without diabetes mellitus had lower fasting insulin concentrations as well as lower HOMA-B and HOMA-IR while fasting glucose
concentrations were not higher. Furthermore, rs7903146-T carriers were characterized by less adiposity (e.g., lower BMI, body weight, waist circumference, and total body fat). After adding BMI or other adiposity measures in the regression model, the association between rs7903146-T and T2D increased, and the association between rs7903146-T and mean fasting insulin concentration attenuated.
Previously, rs7903146-T has been associated with an increased risk of T2D [4-8], as well as with a lower BMI [25]. However, the interrelations of rs7903146 with BMI and T2D are unclear. The rs7903146-T variant has been associated with impaired insulin secretion [16, 30] and impaired beta-cell function [18], while on the other hand rs7903146-T has been associated with higher nocturnal glucose levels [21] and hepatic glucose production [16]. For the association with measures of glucose-insulin metabolism, we observed that rs7903146-T was associated with a lower fasting insulin, lower HOMA-B, and lower HOMA-IR.
However, in contrast to some other studies [40, 41], we did not observe an association between rs7903146-T and fasting glucose concentrations. In line, a Genome-Wide
Association Study on fasting glucose concentration did not observe TCF7L2 among the top hits, suggesting a limited effect size of rs7903146 on fasting glucose concentration [42].
Together with our findings, these results suggest that TCF7L2 is associated with an increased T2D risk by altering pancreatic function, which is reflected by a lower pancreatic insulin
secretion, as has been suggested earlier [14]. The hypothesis of a role of rs7903146 in beta- cell dysfunction is supported by our finding that, among a population of participants with diabetes mellitus, rs7903146-T allele carriers are more likely to use insulin as medication. As there is no evidence in the literature that genetic variation in TCF7L2 (e.g., rs7903146) is related to type-1 diabetes mellitus [43], the higher use of insulin in rs7903146-T carriers might be a reflection of an insulin-dependent T2D phenotype.
In addition to the association of rs7903146-T with a lower BMI [25], rs7903146-T was also associated with other measures of adiposity as total body fat, and waist
circumference. The associations with subcutaneous adipose tissue, visceral adipose tissue, and hepatic triglyceride content showed wide 95% confidence intervals and were not
statistically significant, which is probably due to the lower sample size with data available on these phenotypes. These findings should therefore be studied in greater detail in larger
cohorts. Nevertheless, the consistency in research findings suggest that rs7903146-T affects overall adiposity rather than specific measures of adiposity.
The association between rs7903146-T and the higher risk of having T2D became stronger after adjustment for measures of adiposity. This might indicate that TCF7L2 affects T2D risk not through adiposity, but instead through a separate biological mechanism. As rs7903146-T associates with both lower BMI and lower insulin, this suggests that the risk of developing T2D through obesity-induced insulin resistance is decreased, while the risk of developing T2D through pancreatic dysfunction is higher. This hypothesis is supported by previous findings that genetic variation in TCF7L2 is stronger associated with T2D in non- obese individuals [28, 44, 45], suggesting a biological mechanism different from the metabolic effects of obesity. Nevertheless, how rs7903146 in TCF7L2 affects beta-cell function and adiposity needs to be explored in more detail in future studies. Taken together, these results suggest that the increased risk of T2D in rs7903146-T carriers originates from
beta-cell dysfunction (reflected by lower insulin secretion). Insulin is a key member of the insulin/insulin-like growth factor family. A lower stimulation of growth promoting signaling pathways may result in a lower body weight.
This study has a few strengths and limitations. The strength of the NEO study is the large sample size with the availability of specific adiposity measures. However, the
measurements of subcutaneous and visceral adipose tissue were done in subpopulations and the number of rs7903146-TT carriers is limited in analyses on these phenotypes.
Furthermore, insulin resistance and secretions were estimated using the HOMA-IR and HOMA-B, respectively instead of using the hyperinsulinemic euglycemic clamp. Diagnosis of T2D was self-reported and is combined with individuals with only one fasting glucose measurement above 7 mmol/L. For this reason, there might be some misclassification of the outcome, which could have resulted in findings in the direction of the zero hypothesis.
In conclusion, within the NEO study population, we replicated the previous
associations of rs7903146 with higher risk of prevalent T2D and lower BMI. The observation of lower insulin in carriers of the rs7903146-T allele supports the hypothesis that TCF7L2 affects T2D risk primarily by affecting pancreatic insulin secretion, and not via peripheral insulin resistance. Future studies should elaborate in more detail in the role of TCF7L2, adiposity, and beta-cell function in the onset of T2D.
Acknowledgements
We express our gratitude to all individuals who participate in the Netherlands Epidemiology in Obesity study. We are grateful to all participating general practitioners for inviting eligible participants. We furthermore thank P. van Beelen and all research nurses for collecting the data and P. Noordijk and her team for sample handling and storage and I. de Jonge, MSc for all data management of the NEO study. The NEO study is supported by the participating Departments, the Division and the Board of Directors of the Leiden University Medical Centre, and by the Leiden University, Research Profile Area ‘Vascular and
Regenerative Medicine’. DOMK is supported by the Dutch Science Organization (ZonMW- VENI Grant 916.14.023). DvH was supported by the European Commission funded project HUMAN (Health-2013-INNOVATION-1-602757).
Conflict of Interest Statement
Dr. van Heemst is supported by a grant from the European Commission during the conduct of the study. All other authors declare that they have no conflict of interest.
References
[1] Prunier C, Hocevar BA, Howe PH. Wnt signaling: physiology and pathology. Growth factors (Chur, Switzerland). 2004;22:141-50.
[2] Jin T. Current Understanding on Role of the Wnt Signaling Pathway Effector TCF7L2 in Glucose Homeostasis. Endocrine reviews. 2016;37:254-77.
[3] Cauchi S, Froguel P. TCF7L2 genetic defect and type 2 diabetes. Current diabetes reports.
2008;8:149-55.
[4] Kho AN, Hayes MG, Rasmussen-Torvik L, Pacheco JA, Thompson WK, Armstrong LL, et al. Use of diverse electronic medical record systems to identify genetic risk for type 2 diabetes within a genome-wide association study. Journal of the American Medical Informatics Association : JAMIA. 2012;19:212-8.
[5] Scott LJ, Mohlke KL, Bonnycastle LL, Willer CJ, Li Y, Duren WL, et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science (New York, NY). 2007;316:1341-5.
[6] Sladek R, Rocheleau G, Rung J, Dina C, Shen L, Serre D, et al. A genome-wide association study identifies novel risk loci for type 2 diabetes. Nature. 2007;445:881-5.
[7] Salonen JT, Uimari P, Aalto JM, Pirskanen M, Kaikkonen J, Todorova B, et al. Type 2 diabetes whole-genome association study in four populations: the DiaGen consortium.
American journal of human genetics. 2007;81:338-45.
[8] Steinthorsdottir V, Thorleifsson G, Reynisdottir I, Benediktsson R, Jonsdottir T, Walters GB, et al. A variant in CDKAL1 influences insulin response and risk of type 2 diabetes.
Nature genetics. 2007;39:770-5.
[9] Kong X, Zhang X, Xing X, Zhang B, Hong J, Yang W. The Association of Type 2 Diabetes Loci Identified in Genome-Wide Association Studies with Metabolic Syndrome and Its Components in a Chinese Population with Type 2 Diabetes. PloS one. 2015;10:e0143607.
[10] Yako YY, Madubedube JH, Kengne AP, Erasmus RT, Pillay TS, Matsha TE.
Contribution of ENPP1, TCF7L2, and FTO polymorphisms to type 2 diabetes in mixed ancestry ethnic population of South Africa. African health sciences. 2015;15:1149-60.
[11] Adeyemo AA, Tekola-Ayele F, Doumatey AP, Bentley AR, Chen G, Huang H, et al.
Evaluation of Genome Wide Association Study Associated Type 2 Diabetes Susceptibility Loci in Sub Saharan Africans. Frontiers in genetics. 2015;6:335.
[12] Fall T, Ingelsson E. Genome-wide association studies of obesity and metabolic syndrome. Molecular and cellular endocrinology. 2014;382:740-57.
[13] Cauchi S, El Achhab Y, Choquet H, Dina C, Krempler F, Weitgasser R, et al. TCF7L2 is reproducibly associated with type 2 diabetes in various ethnic groups: a global meta-analysis.
Journal of molecular medicine (Berlin, Germany). 2007;85:777-82.
[14] Helgason A, Palsson S, Thorleifsson G, Grant SF, Emilsson V, Gunnarsdottir S, et al.
Refining the impact of TCF7L2 gene variants on type 2 diabetes and adaptive evolution.
Nature genetics. 2007;39:218-25.
[15] Xia Q, Chesi A, Manduchi E, Johnston BT, Lu S, Leonard ME, et al. The type 2 diabetes presumed causal variant within TCF7L2 resides in an element that controls the expression of ACSL5. Diabetologia. 2016;59:2360-8.
[16] Lyssenko V, Lupi R, Marchetti P, Del Guerra S, Orho-Melander M, Almgren P, et al.
Mechanisms by which common variants in the TCF7L2 gene increase risk of type 2 diabetes.
The Journal of clinical investigation. 2007;117:2155-63.
[17] Le Bacquer O, Kerr-Conte J, Gargani S, Delalleau N, Huyvaert M, Gmyr V, et al.
TCF7L2 rs7903146 impairs islet function and morphology in non-diabetic individuals.
Diabetologia. 2012;55:2677-81.
[18] Florez JC, Jablonski KA, Bayley N, Pollin TI, de Bakker PI, Shuldiner AR, et al.
TCF7L2 polymorphisms and progression to diabetes in the Diabetes Prevention Program.
The New England journal of medicine. 2006;355:241-50.
[19] Pearson ER, Donnelly LA, Kimber C, Whitley A, Doney AS, McCarthy MI, et al.
Variation in TCF7L2 influences therapeutic response to sulfonylureas: a GoDARTs study.
Diabetes. 2007;56:2178-82.
[20] Saxena R, Gianniny L, Burtt NP, Lyssenko V, Giuducci C, Sjogren M, et al. Common single nucleotide polymorphisms in TCF7L2 are reproducibly associated with type 2 diabetes and reduce the insulin response to glucose in nondiabetic individuals. Diabetes.
2006;55:2890-5.
[21] van der Kroef S, Noordam R, Deelen J, Akintola AA, Jansen SW, Postmus I, et al.
Association between the rs7903146 Polymorphism in the TCF7L2 Gene and Parameters Derived with Continuous Glucose Monitoring in Individuals without Diabetes. PloS one.
2016;11:e0149992.
[22] Wegner L, Hussain MS, Pilgaard K, Hansen T, Pedersen O, Vaag A, et al. Impact of TCF7L2 rs7903146 on insulin secretion and action in young and elderly Danish twins. The Journal of clinical endocrinology and metabolism. 2008;93:4013-9.
[23] Westra HJ, Peters MJ, Esko T, Yaghootkar H, Schurmann C, Kettunen J, et al.
Systematic identification of trans eQTLs as putative drivers of known disease associations.
Nature genetics. 2013;45:1238-43.
[24] Karczewska-Kupczewska M, Stefanowicz M, Matulewicz N, Nikolajuk A, Straczkowski M. Wnt signaling genes in adipose tissue and skeletal muscle of humans with different degree of insulin sensitivity. The Journal of clinical endocrinology and metabolism.
2016:jc20161594.
[25] Locke AE, Kahali B, Berndt SI, Justice AE, Pers TH, Day FR, et al. Genetic studies of body mass index yield new insights for obesity biology. Nature. 2015;518:197-206.
[26] Mokdad AH, Ford ES, Bowman BA, Dietz WH, Vinicor F, Bales VS, et al. Prevalence of obesity, diabetes, and obesity-related health risk factors, 2001. Jama. 2003;289:76-9.
[27] Corbin LJ, Richmond RC, Wade KH, Burgess S, Bowden J, Smith GD, et al. Body mass index as a modifiable risk factor for type 2 diabetes: Refining and understanding causal estimates using Mendelian randomisation. Diabetes. 2016.
[28] Timpson NJ, Lindgren CM, Weedon MN, Randall J, Ouwehand WH, Strachan DP, et al.
Adiposity-related heterogeneity in patterns of type 2 diabetes susceptibility observed in genome-wide association data. Diabetes. 2009;58:505-10.
[29] Manning AK, Hivert MF, Scott RA, Grimsby JL, Bouatia-Naji N, Chen H, et al. A genome-wide approach accounting for body mass index identifies genetic variants
influencing fasting glycemic traits and insulin resistance. Nature genetics. 2012;44:659-69.
[30] de Mutsert R, den Heijer M, Rabelink TJ, Smit JW, Romijn JA, Jukema JW, et al. The Netherlands Epidemiology of Obesity (NEO) study: study design and data collection.
European journal of epidemiology. 2013;28:513-23.
[31] Definition and diagnosis of diabetes mellitus and intermediate hyperglycaemia. In:
Federation WHOaID, editor. Report of a WHO/IDF consultation2006. p. 50.
[32] Oxford TUo. HOMA calculator. 2004.
[33] 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;28:412-9.
[34] van der Meer RW, Rijzewijk LJ, de Jong HW, Lamb HJ, Lubberink M, Romijn JA, et al.
Pioglitazone improves cardiac function and alters myocardial substrate metabolism without affecting cardiac triglyceride accumulation and high-energy phosphate metabolism in patients with well-controlled type 2 diabetes mellitus. Circulation. 2009;119:2069-77.
[35] van der Meer RW, Hammer S, Lamb HJ, Frolich M, Diamant M, Rijzewijk LJ, et al.
Effects of short-term high-fat, high-energy diet on hepatic and myocardial triglyceride content in healthy men. The Journal of clinical endocrinology and metabolism.
2008;93:2702-8.
[36] Widya RL, de Mutsert R, den Heijer M, le Cessie S, Rosendaal FR, Jukema JW, et al.
Association between Hepatic Triglyceride Content and Left Ventricular Diastolic Function in a Population-based Cohort: The Netherlands Epidemiology of Obesity Study. Radiology.
2016;279:443-50.
[37] Korn EL, Graubard BI. Epidemiologic studies utilizing surveys: accounting for the sampling design. American journal of public health. 1991;81:1166-73.
[38] Lumley T. Analysis of complex survey samples. 2004.
[39] VWS Mv. Hoeveel mensen hebben overgewicht? . 2013.
[40] Stolerman ES, Manning AK, McAteer JB, Fox CS, Dupuis J, Meigs JB, et al. TCF7L2 variants are associated with increased proinsulin/insulin ratios but not obesity traits in the Framingham Heart Study. Diabetologia. 2009;52:614-20.
[41] Warodomwichit D, Arnett DK, Kabagambe EK, Tsai MY, Hixson JE, Straka RJ, et al.
Polyunsaturated fatty acids modulate the effect of TCF7L2 gene variants on postprandial lipemia. The Journal of nutrition. 2009;139:439-46.
[42] Dupuis J, Langenberg C, Prokopenko I, Saxena R, Soranzo N, Jackson AU, et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nature genetics. 2010;42:105-16.
[43] Ruchat SM, Elks CE, Loos RJ, Vohl MC, Weisnagel SJ, Rankinen T, et al. Evidence of interaction between type 2 diabetes susceptibility genes and dietary fat intake for adiposity and glucose homeostasis-related phenotypes. Journal of nutrigenetics and nutrigenomics.
2009;2:225-34.
[44] Kong X, Xing X, Hong J, Zhang X, Yang W. Genetic variants associated with lean and obese type 2 diabetes in a Han Chinese population: A case-control study. Medicine.
2016;95:e3841.
[45] Al-Safar H, Hassoun A, Almazrouei S, Kamal W, Afandi B, Rais N. Association of the Genetic Polymorphisms in Transcription Factor 7-Like 2 and Peroxisome Proliferator- Activated Receptors- gamma 2 with Type 2 Diabetes Mellitus and is Interaction with Obesity Status in Emirati Population. Journal of diabetes research. 2015;2015:129695.
Appendices
Supplementary Table 1.
Supplementary Table 2.
Supplementary Table 3.
Supplementary Table 4.
FIGURE LEGENDS
Figure 1: Mediation by BMI of the association of rs7903146 with prevalent diabetes, fasting insulin, and HOMA-B
Results were based on analyses weighted towards the BMI distribution of the general population. A) Additive effect of rs7903146-T on T2D. Results presented as odds ratio with 95% confidence interval. B) Additive effect of rs7903146-T on fasting insulin. Results presented as additive effect with 95% confidence interval. C) Additive effect of rs7903146-T on HOMA-B. Results presented as additive effect with 95% confidence interval. Model 1 adjusted for age, sex, and principal components. Mode 2 adjusted for age, sex, body mass index, and principal components. Participants using glucose-lowering medication were excluded from the analyses presented in Figure 1B and 1C. The asterisk represents a p-value
< 0.05.
