Causal association between inflammatory bowel disease and type 2 diabetes: a bidirectional two-sample Mendelian randomization study

doi:10.21203/rs.3.rs-3340188/v1

Download PDF

Research Article

Causal association between inflammatory bowel disease and type 2 diabetes: a bidirectional two-sample Mendelian randomization study

https://doi.org/10.21203/rs.3.rs-3340188/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Background

Previous observational studies have shown a paradoxical association between inflammatory bowel disease (IBD) and type 2 diabetes (T2DM), and the causal relationship between the two is unclear. Using a bidirectional two-sample Mendelian randomization (MR), we explored the causal connection between IBD and T2DM.

Methods

We obtained the datasets of IBD, UC, CD and T2DM from IEU genome-wide association study (GWAS) summary statistics and extracted genetic variants as instrumental variables for bidirectional two-sample MR. The main MR analysis methods are Inverse-variance weighted. Mendelian randomization-Egger (MR-Egger) and the weighted median as a complementary MR analysis method. Finally, heterogeneity tests, horizontal multiple validity tests, and the leave-one-out sensitivity analysis were used to evaluate the robustness of the study results.

Results

MR analysis results showed no causal connection between IBD (IVW OR = 1.017, 95%CI 0.981–1.053, p = 0.361), ulcerative colitis (UC) (IVW OR = 0.997, 95%CI 0.964–1.031, p = 0.864), Crohn’s disease (CD) (IVW OR = 1.018, 95%CI 0.986–1.052, p = 0.27) and T2DM. In addition, the results of the reverse MR analysis revealed no proof to support the risk of T2DM with the development of IBD (IVW OR = 1.049, 95%CI 0.898–1.225, p = 0.544), UC (IVW OR = 0.945, 95%CI 0.801–1.115, p = 0.501), and CD (IVW OR = 1.085, 95%CI 0.928–1.27, p = 0.307). Similar results were obtained for the complementary MR analysis method. The SNPs used in this study exhibited no conspicuous horizontal pleiotropy, and sensitivity analysis showed the robustness of the findings.

Conclusion

Our bidirectional two-sample MR results showed no evidence to support a causal connection between IBD, UC, CD and T2DM in the European population.

Inflammatory bowel disease

Ulcerative colitis

Crohn’s disease

Type 2 diabetes

Mendelian randomization

IBD is a chronic inflammatory disease of the intestine, consisting mainly of UC and CD, which presents with abdominal pain, watery or bloody diarrhea, weight loss, and extraintestinal symptoms. Endoscopy of UC of varying severity may show erythema, friability, spontaneous bleeding, erosion and ulceration from the distal to the proximal colon, while the different presentation of CD is a jumpy cobblestone-like lesion of the colon and ileum. ^{[1, 2]} As IBD disease progresses there may be an increased risk of anxiety and depression, as well as a serious psychosocial and economic burden for patients.^{[3, 4]} Globally, the prevalence and prevalence of IBD continues to increase progressively. ^{[5, 6]} Diabetes mellitus is a common chronic endocrine disease. Additionally, according to research, there will be 693 million diabetics globally by 2045, with the incidence of the disease expected to continue rising, causing a serious medical burden on society. ^{[7, 8]} Some epidemiology suggests a connection between IBD and DM. Study shows that IBD patients have a higher chance of developing DM.^[9] Moreover, different IBD subtypes were revealed to have different diabetes risk factors in IBD patients. Study finds UC patients associated with higher risk of diabetes, but not CD patients.^{[10, 11]} On the contrary, a retrospective cohort study in Korea showed that only CD patients had a noticeably elevated risk of DM.^[12] DM may increase the serious risk of hospitalization, sepsis, and infection in IBD patients, increasing health care utilization and affecting patients' quality of life.^{[13, 14]} Regarding the association between IBD and T2DM, there is ambivalence between the results of current observational studies.

The Mendelian randomization research approach employs genetic loci as instrumental variables (IV) to investigate the causal effects of exposure on outcome, which can effectively avoid confounding factors in observational studies and effects due to reverse causality.^[15] Therefore, this study investigated the causal connection between IBD and T2DM using a bidirectional two-sample MR based on GWAS public database.

2.1 Study design

The study was a bidirectional two-sample MR, and genetic instruments for both exposure and outcome were extracted from the IEU Open GWAS database (https://gwas.mrcieu.ac.uk/). IBD (including UC and CD) was first used as an exposure factor to assess its causal effect on T2DM, and then T2DM was used as an exposure variable to investigate its reverse causal effect on IBD. The MR analysis was consistent with three assumptions: (1) the association assumption: Exposure and genetic variants are strongly connected; (2) the independence assumption: genetic variants cannot be connected to exposure or outcome related confounders; (3) the exclusivity assumption: IV can only influence outcome variables through exposure factors.^[16]

2.2 Data source

The IBD, UC, CD, and T2DM datasets were searched separately in the IEU GWAS public database, and all participants were European. The IBD, UC, and CD datasets were derived from the International IBD Genetics Consortium (IIBDGC) GWAS across pedigrees, based on histopathology, endoscopy, and imaging as diagnostic criteria for IBD and its subtypes.^[17] The IBD dataset included 31,665 IBD cases and 33,977 healthy controls, the UC dataset included 13,768 UC cases and 33,977 healthy controls, and the CD dataset included 17,897 CD cases and 33,977 healthy controls. The T2DM dataset was derived from a meta-analysis created by Bonas-Guarch et al, which included 12,931 T2DM cases and 57,196 healthy controls.^[18] The dataset was obtained from the publicly available GWAS database and therefore did not require additional ethical approval.

2.3 Selection of instrumental variables

SNPs that complied with the significance criteria (P < 5×10^− 8) were selected as initial variables in the exposure factor GWAS, and then SNPs with linkage disequilibrium (LD) were excluded based on (r² < 0.01, window size > 10000 kb), and significant and independent SNPs were selected as IVs. Look for second phenotypes of SNPs in PhenoScanner V2 to exclude confounding (http://www.phenoscanner.medschl.cam.ac.uk/).^[19] Identify SNPs corresponding to outcome data and without significance criteria (P < 5×10^− 8). The F-statistic was used as a measure of the strength of IV, and F-statistic < 10 was referred to as a weak instrumental variable, and the corresponding SNPs were excluded from the study.^[20]

2.4 Mendelian randomization

Using inverse-variance weighted as the main method for MR analysis ^[21], Mendelian randomization–Egger (MR-Egger) method^[22] and the weighted median method^[23] were also used to calculate the causative relationships between IBD, UC, CD and T2DM to improve the reliability of the study's findings. Odds ratios (ORs) and 95% confidence intervals (CIs) were used to evaluate the potential causal connection between IBD, UC, CD and T2DM. The "TwoSampleMR" package of R 4.3.1 was used to carry out the complete analytic process.

2.5 Pleiotropy and sensitivity analysis

A heterogeneity test was conducted using Cochrane's Q statistic, primarily to examine variations in individual IVs. The estimation of MR was done using a random effects model where there was significant heterogeneity (p < 0.05). The F-statistic(F = R²/(1-R²) * (n-k-1)/k) was utilized to evaluate genetic instrument strength. MR-Egger analysis is used to detect horizontal pleiotropy. The intercept term can be used to indicate whether there is horizontal pleiotropy among the IVs, and if it is almost exactly 0, the less likely it is to be affected by potential pleiotropy.^[24] “The leave-one-out” sensitivity analysis calculated the combined effect of the remaining SNPs after eliminating SNPs one by one. After each SNP was eliminated individually, the combined effect of the remaining SNPs was determined using sensitivity analysis.

3.1 Selected single-nucleotide polymorphisms in Mendelian randomization

134, 88, 122, and 16 SNPs were selected as IVs for IBD, UC, CD, and T2DM, respectively, based on significance and independence criteria. The F-statistic values for IBD, UC, CD, and T2DM were significantly greater than 10, and the range of F-statistic values corresponding to a single SNP was 97.7 to 3614.1, 110 to 6086.4, 114.4 to 3044.5, and 279.4 to 3862.6, respectively. Thus, suggesting that the causal association between exposure and outcome is less likely to be affected by weak instrumental variable bias. The effect of confounding factors was excluded by PhenoScanner V2 to satisfy the exclusion hypothesis and to exclude SNPs with significance criteria in the outcome data. After a rigorous screening process when the outcome factor was T2DM, 122, 78 and 108 SNPs were finally chosen to be IVs for IBD, UC and CD, respectively; when the outcome factor was IBD, 6 SNPs were finally chosen to be IVs for T2DM; when the outcome factor was UC, 7 SNPs were finally chosen to be IVs for T2DM; and when the outcome factor was CD, 5 SNPs were finally chosen to be IVs for T2DM.

3.2 Causal effect of IBD, UC, CD on T2DM

The IVW method showed no causal effect of IBD on T2DM (OR = 1.019, 95% CI 0.984–1.056, p = 0.290), consistent with the MR-Egger method and the Weighted median method (Table 1; Fig. 1A). Supplementary Figs. 1 and 2 show the scatter plot and funnel plot of IBD, UC, CD and T2DM associated SNPs, respectively. MR Egger regression tested for pleiotropy and the results indicated that horizontal pleiotropy of SNPs was less likely to affect causality in IBD (Table 2). Cochran Q and I² values indicated significant heterogeneity between IV estimates, so random effect model is used to analyze ICD, CD and T2DM (Table 2). The results of the Cochran Q test suggested that there was no heterogeneity between the IV estimates and therefore a fixed effects model was used to analyze UC and T2DM (Q = 95.722, p = 0.073). The results of the "the leave-one-out" sensitivity analysis showed that there was no significant effect of any of the SNPs on the causal relationship between IBD, UC, CD and T2DM (Supplementary Fig. 3).

Table 1

Causal effects estimates of MR analyse.
Exposure	Outcome	MR methods	OR	95% CI lower	95% CI upper	p Value
IBD	T2DM	IVW	1.017	0.981	1.053	0.361
		MR Egger	1.005	0.924	1.093	0.907
		Weighted median	1.026	0.978	1.077	0.290
UC	T2DM	IVW	0.997	0.964	1.031	0.864
		MR Egger	1.035	0.944	1.135	0.468
		Weighted median	1.017	0.965	1.072	0.518
CD	T2DM	IVW	1.018	0.986	1.052	0.270
		MR Egger	0.994	0.911	1.084	0.887
		Weighted median	1.018	0.977	1.060	0.391
T2DM	IBD	IVW	1.049	0.898	1.225	0.544
		MR Egger	1.035	0.708	1.512	0.869
		Weighted median	1.044	0.982	1.109	0.169
T2DM	UC	IVW	0.945	0.801	1.115	0.501
		MR Egger	1.029	0.690	1.534	0.894
		Weighted median	0.976	0.909	1.047	0.496
T2DM	CD	IVW	1.085	0.928	1.270	0.307
		MR Egger	1.101	0.756	1.605	0.650
		Weighted median	1.068	0.994	1.148	0.072
MR, Mendelian randomization; IBD, inflammatory bowel disease; UC, ulcerative colitis; CD, Crohn’s disease; T2DM, type 2 diabetes. SNPs, Number of single nucleotide polymorphism.

Table 2

Heterogeneity and pleiotropy test of MR analyse.
Exposure	Outcome	SNPs	Heterogeneity test				MR Egger pleiotropy test
Exposure	Outcome	SNPs	Q (MR Egger)	p-value (MR Egger)	Q (IVW)	p-value (IVW)	Intercept	p-value
IBD	T2DM	121	161.813	0.006	161.935	0.006	0.001	0.765
UC	T2DM	79	95.722	0.073	96.654	0.075	-0.005	0.389
CD	T2DM	108	161.246	0.0004	161.793	0.0005	0.004	0.550
T2DM	IBD	6	54.167	4.86E-11	54.255	1.86E-10	0.003	0.940
T2DM	UC	7	47.777	3.94E-09	49.862	5.01E-09	-0.017	0.660
T2DM	CD	5	26.954	6.02E-06	27.021	1.97E-05	-0.003	0.936
MR, Mendelian randomization; SNPs, Number of single nucleotide polymorphism, IBD, inflammatory bowel disease; UC, ulcerative colitis; CD, Crohn’s disease; T2DM, type 2 diabetes.

3.3 Causal effect of T2DM on IBD, UC, CD

There was no association between T2DM and IBD (IVW OR = 1.049, 95%CI 0.898–1.225, p = 0.544), UC (IVW OR = 0.945, 95%CI 0.801–1.115, p = 0.501), CD (IVW OR = 1.085, 95%CI 0.928–1.27, p = 0.307), according to a reverse MR analysis to determine the causative influence of T2DM on IBD, UC, CD (Table 1; Fig. 1). The MR-Egger method and the Weighted median method showed no causal effect of T2DM on IBD, UC, CD, consistent with the IVW method. Supplementary Figs. 1 and 2 shows the scatter plot and funnel plot of SNPs associated with T2DM and IBD, UC, CD, respectively. MR Egger regression tested for pleiotropy and the results suggest that horizontal pleiotropy of SNPs is unlikely to affect causality in T2DM (Table 2). Significant variability in the IV estimations was shown by the Cochran Q and I² values, which a random effects model was used to investigate (Table 2). The results of the "the leave-one-out" sensitivity analysis revealed that none of the SNPs had a substantial impact on the causal relationship between T2DM and IBD, UC, CD (Supplementary Fig. 3).

In our study, we used bidirectional two-sample MR analysis and GWAS public database for the first time to assess the causative link between IBD, UC, CD and T2DM. Our findings suggest that there is not enough proof to support a causal role of IBD, UC, and CD in the risk of T2DM in the European population. Additionally, there is also no inverse causal connection between them. The results of our MR analysis using three different methods were consistent, as well as the results of the sensitivity analysis, both of which indicated the robustness of the results.

Our findings differ from those of earlier observational research, which is a contradiction. A Danish cohort study including 736,072 individuals showed a 50% increased risk of T2DM in patients with IBD contrasted with the general population, and a similar risk of T2DM in UC and CD patients had a similar risk of T2DM. ^[9] The distinction is that certain research revealed a different risk of acquiring T2DM in individuals with UC and CD. Dregan et al^[10] found that UC patients had a higher risk of developing DM than those with CD. Similar findings were made by Kappelman et al ^[11], who discovered that among juvenile IBD patients, only individuals with UC were linked to a higher risk of DM, with no discernible relationship between patients with CD and an increased incidence of DM. In contrast, a retrospective cohort research conducted in Korea revealed that patients with CD had a considerably higher risk of acquiring DM, compared to those with UC. ^[12] Observational studies are susceptible to confounding factors such as socioeconomic, lifestyle, and dietary patterns, so the previously observed significant associations between IBD, UC, CD, and T2DM may be biased and do not indicate a causal relationship. These confounding variables and the impact of reverse causality may both be avoided in MR research, and the current study shows no causal or reverse causality between genetically predicted IBD, UC, CD, and T2DM.

In 141,170 T2DM people, UK cohort research reported a prevalence of IBD of 37.7/100,000 person-years. The use of DPP-4 inhibitors was substantially related to a higher risk of UC, but no significant connection was seen with the risk of CD.^[25] In contrast, animal studies have found that metformin reduces intestinal inflammation by inhibiting STAT3 phosphorylation and promoting STAT5 phosphorylation, and that metformin can be used to prevent the development of IBD .^[26] According to Tseng et al^[27], metformin may operate by decreasing intestinal inflammation and improving insulin resistance in people with T2DM, considerably lowering their chance of developing IBD. The use of drugs during the treatment of diabetes may have different effects on the outcome of IBD, so the use of different therapeutic drugs may be linked to the higher risk of IBD in DM individuals, which remains to be confirmed by further studies. Allin et al ^[28] studied the risk of T2DM in individuals with bowel resected IBD and found that individuals with small bowel resected CD had a reduced risk of T2DM, however, individuals with large bowel resected UC had a considerably higher chance of developing T2DM. Glucocorticoids are one of the most important treatments for patients with moderate to severe IBD, however, glucocorticoids can reduce insulin sensitivity, which can lead to the development of diabetes .^[29] Thus hormonal therapy in IBD patients may increase the risk of developing T2DM, while surgical treatment also leads to different risks of developing T2DM in different subtypes. Observational studies are vulnerable to inexact risk relationships arising from the confounding factors described above.

IBD patients have a higher risk of malnutrition due to reduced dietary intake, impaired absorption due to intestinal inflammation, and loss of nutrients due to diarrhea, especially in patients with CD .^{[30, 31]} This goes against the risk factors for DM such as obesity and high-fat diets. Metabolic syndrome refers to a range of risk factors such as obesity, insulin resistance, dyslipidemia and hypertension as metabolic factors that increase the risk of T2MD. Excess circulating fatty acids in the body lead to insulin resistance emergence, and the prevalence of metabolic syndrome increases significantly with the increase of insulin resistance.^[32] Insulin resistance is an early manifestation of diabetes. DOĞAN et al^[33] assessed insulin resistance status by HOMA-IR index in 95 patients with IBD and found that HOMA-IR levels were higher in UC patients than in controls, with no significant difference in CD patients. In addition, several studies have demonstrated that the prevalence of metabolic syndrome in IBD individuals is not significantly different from the general population.^[34] Insulin resistance as well as metabolic syndrome are risk factors for T2DM, but the underlying pathogenesis needs to be further investigated in patients with IBD.

Bile acids and bile acid metabolism are closely related to the development of IBD and DM. Labbe et al^[35] found that the abundance of bile salt hydrolase gene (bsh), 7 alpha-dehydroxylase gene (adh) and 7-alpha hydroxysteroid dehydrogenase gene (hsdh) were significantly reduced in the intestine of both UC and T2DM patients, while adh and hsdh abundance were increased in CD patients, using publicly available macrogenomic information analysis. Farnesoid X Receptor (FXR) suppresses intestinal inflammation and bacterial overgrowth by regulating genes in IBD, and BAs regulate blood glucose by reducing hepatic gluconeogenesis and enhancing glycogen synthesis, or G-protein coupled receptor (TGR-5) signaling increases GLP-1 secretion in T2DM. Intestinal microbiota dysbiosis and increased permeability of the intestinal mucosa in patients with T2DM.^[36] Intestinal inflammation and microbiome alterations in IBD patients lead to intestinal mucosa barrier dysfunction and disrupt the homeostasis of microbial-derived metabolites.^[37] Both T2DM and IBD suffer from an imbalance in the gut microbiome and consequent destruction of the intestinal mucosa. At the same time, the chronic inflammatory process releases various cytokines leading to dysregulation of the gut microbiota, in which TGF-β, TNFα, and NFKB play important roles. Patients with IBD have higher levels of TNF-α expression in their mucosa, TGF-β upregulates Th17-related cytokines as well as IL-6 and NFKB upregulates chemokines and cytokines, and these pathways together contribute to the development of IBD; in T2DM, TGFβ induces β-cell dysfunction, NFKB increases the risk of developing T2DM in obese people, and TNF-α promotes the progression of diabetes through effects on cell adhesion molecules and thus promote the progression of diabetes.^[38] These inflammatory pathways are independent of each disease and at the same time crosstalk with each other. In summary, chronic inflammatory processes, bile acids, and gut microbiota may be shared potential contributors to IBD and T2DM, which have been found to coexist in observational clinical studies, but do not constitute a causal relationship between the two factors.

Our study still has some limitations. To begin with, the study was conducted on a population of European ancestry, a finding that may have limitations for populations of other races. Secondly, we only stratified the study on IBD and its subtypes, and we were unable to stratify the study on exposure and outcome according to different factors such as age and gender due to the limitation of GWAS data. Finally, there was significant heterogeneity in the articles, but the robustness of the results was verified by multiple validity tests and sensitivity analysis.

Our bidirectional two-sample MR results showed no evidence to support a causal connection between IBD, UC, CD and T2DM in the European population. Treatment of IBD individuals may not decrease the risk of developing T2DM and vice versa. To further investigate the possible association between IBD and its subtypes and T2DM, more prospective studies and MR analysis are required.

Acknowledgments

Thanks to the Gansu Province Clinical Research Center for Digestive Diseases, the First Hospital of Lanzhou University, for their support.

Ethics approval and consent to participate

The dataset was obtained from the publicly available GWAS database and therefore did not require.

Consent for publication

All the authors of the article agreed to publish it.

Availability of data and materials

The datasets generated and analysed during the current study are available in the GWAS database

(https://gwas.mrcieu.ac.uk/).

Competing interests

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding

The article did not receive financial support.

Authors' contributions

MjY and YjY acquisition of data, analysis and interpretation of data, drafting the article, final approval; MZ, HC and JOY interpretation of data, revising the article, final approval; YZ and YpW revising the article, final approval; QhG conception and design of the study, critical revision, resolve discrepancies, final approval.

Flynn S, Eisenstein S. Inflammatory Bowel Disease Presentation and Diagnosis. Surg Clin North Am, 2019. 99(6): p. 1051-1062.
Kaenkumchorn T, Wahbeh G. Ulcerative Colitis: Making the Diagnosis. Gastroenterol Clin North Am, 2020. 49(4): p. 655-669.
Naegeli AN, Balkaran BL, Shan M, et al. The impact of symptom severity on the humanistic and economic burden of inflammatory bowel disease: a real-world data linkage study. Curr Med Res Opin, 2022. 38(4): p. 541-551.
Halloran J, McDermott B, Ewais T, et al. Psychosocial burden of inflammatory bowel disease in adolescents and young adults. Intern Med J, 2021. 51(12): p. 2027-2033.
Park J, Cheon JH. Incidence and Prevalence of Inflammatory Bowel Disease across Asia. Yonsei Med J, 2021. 62(2): p. 99-108.
Zhao M, Gönczi L, Lakatos PL, et al. The Burden of Inflammatory Bowel Disease in Europe in 2020. J Crohns Colitis, 2021. 15(9): p. 1573-1587.
Cho NH, Shaw JE, Karuranga S, et al. IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045. Diabetes Res Clin Pract, 2018. 138: p. 271-281.
Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract, 2010. 87(1): p. 4-14.
Jess T, Jensen BW, Andersson M, et al. Inflammatory Bowel Diseases Increase Risk of Type 2 Diabetes in a Nationwide Cohort Study. Clin Gastroenterol Hepatol, 2020. 18(4): p. 881-888 e1.
Dregan A, Charlton J, Chowienczyk P, et al. Chronic inflammatory disorders and risk of type 2 diabetes mellitus, coronary heart disease, and stroke: a population-based cohort study. Circulation, 2014. 130(10): p. 837-44.
Kappelman MD, Galanko JA, Porter CQ, et al. Association of paediatric inflammatory bowel disease with other immune-mediated diseases. Arch Dis Child, 2011. 96(11): p. 1042-6.
Kang EA, Han K, Chun J, et al. Increased Risk of Diabetes in Inflammatory Bowel Disease Patients: A Nationwide Population-based Study in Korea. J Clin Med. 2019 Mar 11;8(3):343.
Fuschillo G, Celentano V, Rottoli M, et al. Influence of diabetes mellitus on inflammatory bowel disease course and treatment outcomes. A systematic review with meta-analysis. Dig Liver Dis, 2022 Sep 1:S1590-8658(22)00645-4.
Din H, Anderson AJ, Ramos Rivers C, et al. Disease Characteristics and Severity in Patients With Inflammatory Bowel Disease With Coexistent Diabetes Mellitus. Inflamm Bowel Dis, 2020. 26(9): p. 1436-1442.
Emdin CA, Khera AV, Kathiresan S. Mendelian Randomization. JAMA. 2017 Nov 21;318(19):1925-1926.
Sekula P, Del Greco M F, Pattaro C, et al. Mendelian Randomization as an Approach to Assess Causality Using Observational Data. J Am Soc Nephrol, 2016. 27(11): p. 3253-3265.
Liu JZ, van Sommeren S, Huang H, et al. Association analyses identify 38 susceptibility loci for inflammatory bowel disease and highlight shared genetic risk across populations. Nat Genet, 2015. 47(9): p. 979-986.
Bonàs-Guarch S, Guindo-Martínez M, Miguel-Escalada I, et al. Re-analysis of public genetic data reveals a rare X-chromosomal variant associated with type 2 diabetes. Nat Commun, 2018. 9(1): p. 321.
Kamat MA, Blackshaw JA, Young R, et al. PhenoScanner V2: an expanded tool for searching human genotype-phenotype associations. Bioinformatics, 2019. 35(22): p. 4851-4853.
Burgess S, Small DS, Thompson SG. A review of instrumental variable estimators for Mendelian randomization. Stat Methods Med Res, 2017. 26(5): p. 2333-2355.
Burgess S, Butterworth A, Thompson SG. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet Epidemiol, 2013. 37(7): p. 658-65.
Bowden J, Davey Smith G, Burgess S. Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression. Int J Epidemiol, 2015. 44(2): p. 512-25.
Bowden J, Davey Smith G, Haycock PC, et al. Consistent Estimation in Mendelian Randomization with Some Invalid Instruments Using a Weighted Median Estimator. Genet Epidemiol, 2016. 40(4): p. 304-14.
Burgess S, Thompson SG. Interpreting findings from Mendelian randomization using the MR-Egger method. Eur J Epidemiol, 2017. 32(5): p. 377-389.
Abrahami D, Douros A, Yin H, et al. Dipeptidyl peptidase-4 inhibitors and incidence of inflammatory bowel disease among patients with type 2 diabetes: population based cohort study. BMJ, 2018. 360: p. k872.
Lee SY, Lee SH, Yang EJ, et al. Metformin Ameliorates Inflammatory Bowel Disease by Suppression of the STAT3 Signaling Pathway and Regulation of the between Th17/Treg Balance. Plos One, 2015. 10(9) :e0135858.
Tseng CH. Metformin Use Is Associated with a Lower Risk of Inflammatory Bowel Disease in Patients with Type 2 Diabetes Mellitus. J Crohns Colitis, 2021. 15(1): p. 64-73.
Allin KH, Agrawal M, Iversen AT, et al. The Risk of Type 2 Diabetes in Patients With Inflammatory Bowel Disease After Bowel Resections: A Nationwide Cohort Study. Gastro Hep Advances, 2022. 1(5): p. 777-784.
Golubic R, Caleyachetty R, Barber TM, et al. Glucocorticoid-induced hyperglycaemia and diabetes: Call for action. Diabet Med, 2022. 39(8): p. e14843.
Bischoff SC, Escher J, Hébuterne X, et al. ESPEN practical guideline: Clinical Nutrition in inflammatory bowel disease. Clin Nutr, 2020. 39(3): p. 632-653.
Cucinotta U, Romano C, Dipasquale V. Diet and Nutrition in Pediatric Inflammatory Bowel Diseases. Nutrients, 2021. 13(2) :655.
Eckel R H, Grundy S M, Zimmet P Z. The metabolic syndrome. Lancet, 2005. 365(9468): p. 1415-28.
Doğan AN, Kahraman R, Akar T. Evaluation of insulin resistance and beta cell activity in patients with inflammatory bowel disease. Eur Rev Med Pharmacol Sci. 2022 Jun;26(11):3989-3994.
Nagahori M, Hyun SB, Totsuka T, et al. Prevalence of metabolic syndrome is comparable between inflammatory bowel disease patients and the general population. J Gastroenterol, 2010. 45(10): p. 1008-13.
Labbé A, Ganopolsky JG, Martoni CJ, et al. Bacterial bile metabolising gene abundance in Crohn's, ulcerative colitis and type 2 diabetes metagenomes. PLoS One, 2014. 9(12): p. e115175.
Yang G, Wei J, Liu P, et al. Role of the gut microbiota in type 2 diabetes and related diseases. Metabolism, 2021. 117: p. 154712.
Verdugo-Meza A, Ye J, Dadlani H, et al. Connecting the Dots Between Inflammatory Bowel Disease and Metabolic Syndrome: A Focus on Gut-Derived Metabolites. Nutrients, 2020. 12(5) :1434.
Jurjus A, Eid A, Al Kattar S, et al. Inflammatory bowel disease, colorectal cancer and type 2 diabetes mellitus: The links. BBA Clin, 2016. 5: p. 16-24.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Causal association between inflammatory bowel disease and type 2 diabetes: a bidirectional two-sample Mendelian randomization study

Status:

Version 1

Abstract

Background

Methods

Results

Conclusion

Figures

1 Introduction

2 Materials and methods

2.1 Study design

2.2 Data source

2.3 Selection of instrumental variables

2.4 Mendelian randomization

2.5 Pleiotropy and sensitivity analysis

3 Results

3.1 Selected single-nucleotide polymorphisms in Mendelian randomization

3.2 Causal effect of IBD, UC, CD on T2DM

3.3 Causal effect of T2DM on IBD, UC, CD

4 Discussion

5 Conclusion

Declarations

References

Additional Declarations

Supplementary Files

Status:

Version 1