Abstract
Inflammatory bowel disease (IBD) and periodontitis are reported to be closely associated; however, whether there is a causal association between them remains unclear. To explore the existence of this causality, this study applied a bidirectional two-sample Mendelian randomization (MR). The genetic variants were obtained from the summary statistics of genome-wide association studies of IBD, including its subtypes CD and UC, and periodontitis. 175, 148, 113, and six single-nucleotide polymorphisms were selected as instrumental variables for IBD, CD, UC, and periodontitis, respectively. In MR analysis, random-effects inverse-variance weighted was used as the primary method, and weighted median and MR Egger regression were applied as the complementary method. A series of sensitivity analyses were also conducted to ensure the reliability of the results. None of these analyses found a significant effect of genetically proxied IBD and its subtypes on periodontitis, and vice versa. Subsequent sensitivity analyses did not detect any horizontal pleiotropy and heterogeneity. Caution should be exerted when it comes to clinical relevance and further studies are needed to clarify the relationship between IBD and periodontitis.
Similar content being viewed by others
Introduction
Inflammatory bowel disease (IBD), mainly comprising Crohn’s disease (CD) and ulcerative colitis (UC), is a chronic, relapsing, and non-specific inflammatory disorder of the intestine1. CD usually affects the entire gastrointestinal tract and leads to transmural inflammation, whereas UC mainly affects the colon and rectum, causing inflammation restricted to the mucosal layer2. The suggested main cause of IBD lies in the interaction between potential pathogenic bacteria and the immune system of the host, which can be influenced by genetic and environmental factors3. In addition to intestinal inflammation, IBD patients also suffer extraintestinal manifestations among which periodontitis is a frequently reported oral manifestation4.
Periodontitis is a chronic inflammatory disease of periodontium, a collective term for the tissues that surround and support the dentition (gingiva, periodontal ligament, and alveolar bone), with an infectious origin5. The features of periodontitis include clinical attachment loss, alveolar bone absorption, periodontal pocketing, gingiva bleeding, and finally tooth loss, compromising mastication, esthetics, and life quality6. The pathogenesis of periodontitis has not been completely understood yet; however, a complex combination of genetic and environmental factors and dysregulated host immune response to oral microbes is widely accepted as the cause7. Furthermore, host genetic variants can affect the immune response to pathogens, increase susceptibility to the colonization of specific bacteria, and affect the threshold of dysbiosis as opposed to self-resolution in response to an acute change8.
Several lines of evidence have suggested the relevance between IBD and periodontitis. A higher prevalence of periodontitis was observed in patients with IBD than in controls in a case–control study (37.5% vs. 19.2%)9. Furthermore, periodontitis in patients with IBD presented more severe symptoms than in controls10. A previous meta-analysis, including eight case–control studies and one longitudinal cohort study, showed clear evidence for a higher prevalence of periodontitis in patients with IBD11. However, no evidence attributes this association to causality. From a medical and therapeutic standpoint, causality has more clinical value than mere association. As an example, if a causal effect of IBD on periodontitis does exist, the treatment of IBD may thus contribute to the improvement of periodontitis in IBD patients. Due to reverse causality, measurement error, and underlying bias, observational studies have limitations in causality inference12. Despite being adjusted, confounders, such as age, sex, smoking, and medications, may still interfere with meta-analyses and lead to biases. Randomized controlled trials (RCTs) are considered the gold standard for causality inference; however, RCTs on the causality between IBD and periodontitis in humans have so far not been conducted due to practicality, cost, and ethical considerations. Sufficient evidence to help understand whether IBD can cause periodontitis is still lacking.
Mendelian randomization (MR) provides a powerful strategy for causality inference. It achieves a randomization design similar to the one in RCTs through randomization inherent in gamete formation and therefore is not subject to confounding or reverse-causality bias13. In MR analysis, if genetic variants, in this case, single nucleotide polymorphisms (SNPs), associated with exposure link to a greater risk of outcome, a causal effect of the genetically proxied exposure on the outcome can be established. Taking advantage of published large-scale genome-wide association studies with no sample overlap, a two-sample MR analysis has great potential in causality inference with high statistical power14.
In this study, we used a two-sample MR framework to detect the potential causal association between IBD and periodontitis. The results can offer robust evidence in exploring the nature of the association between these two conditions.
Materials and methods
Study design
We applied two-sample MR analysis using instrumental variable (IV) estimation. IVs must meet the following three assumptions (Fig. 1): (1) IVs robustly associate with exposure (relevance). (2) IVs do not associate with confounders (exchangeability). (3) IVs affect the outcome only through exposure rather than any other ways (exclusion restriction)15. This study was approved by the human research ethics committee in the Shanxi Medical University School and Hospital of Stomatology (2022SLL029). This study adhered to the rigorous Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for MR papers16.
Data source
All datasets in this study are publicly available and show no sample overlap. Summary statistics of periodontitis were obtained from a genome-wide association study (GWAS) of European studies of the Gene-Lifestyle Interactions in Dental Endpoints (GLIDE) consortium, including 17,353 clinical periodontitis cases and 28,210 control cases17. Periodontitis cases were defined by the Centers for Disease Control and Prevention/American Academy of Periodontology or the Community Periodontal Index case definition or via study participant reports of the diagnosis of periodontitis17. GWAS for the periodontitis trait applied a mixed logistic model and adjusted for age, sex, and genetic principal components. Genotyping was performed on commercially available arrays (including the Affymetrix Genome-Wide Human SNP array, Illumina Human610 Quadv1_B, and Affymetrix UK BiLEVE Axiom array) and standard quality control checks were performed before imputation in this GWAS meta-analysis. Summary statistics for IBD (25,042 cases; 34,915 controls), UC (12,366 cases; 33,609 controls), and CD (12,194 cases; 28,072 controls) were derived from a GWAS of European participants18. The diagnosis of IBD was done using accepted endoscopic, histopathological, and radiological criteria. The study performed a fixed-effects meta-analysis; IBD and control cases in this GWAS were genotyped on the Human Core Exome v12.1. The details of the quality control method are provided in the original manuscript. Both data sources are briefly described in Table S1. The included GWASs obtained ethical review board approval and informed consent as described in the respective original manuscript.
Instrument variants selection
We performed a set of quality control techniques to fulfill the first MR assumption. SNPs, acting as IVs in the present MR analysis, were selected at genome-wide significance (P < 5 × 10−8 for IBD and P < 5 × 10−6 for periodontitis), and palindromic variants were removed. The selected SNP unavailable in the outcome datasets was also removed. Then, we performed linkage disequilibrium (LD) clumping to ensure that SNPs were independent (clumping r2 cut-off = 0.001 and clumping window = 10,000 kb). SNP with an r2 larger than the stated threshold was retained. The strength of IVs was further examined using the F-statistic. SNPs with F-statistics < 10 were removed to avoid weak instrument bias for further verifying the relevance assumption. We also ruled out potential pleiotropy by manually searching all IVs in the PhenoScanner for detecting previously reported confounders, including smoking, drinking, and body mass index. We excluded IVs associated with these confounders from the following analysis.
MR analysis
We harmonized summary statistics results to ensure that the effect allele of each SNP was consistent between exposure and outcome and to prohibit strand mismatch. This two-sample MR applied the random-effects inverse-variance weighted (IVW)19, the weighted median20 and the MR Egger regression21 to evaluate the causal effects of exposures on outcomes. With great statistical power and sensitivity to directional horizontal pleiotropy, the IVW method was treated as a fundamental analysis method. The result of the IVW method is robust if all IVs are valid or if any horizontal pleiotropy is balanced22.
Sensitivity analysis
The violations of the second and third assumptions can occur through horizontal pleiotropy (the genetic variant affects the exposure and outcome but via different pathways). Sensitivity analysis can be used to uncover the possible violations of these assumptions. The Egger regression intercept was used to assess the presence of horizontal pleiotropy21. Intercept centered at the origin with a 95% confidence interval including the null reveals the absence of pleiotropy. The MR pleiotropy residual sum and outlier test were also used as complementary analyses to indicate the presence of pleiotropy23. We used the IVW method and MR Egger regression to detect heterogeneity, which was qualified through Cochrane's Q statistic (P < 0.1 indicates heterogeneity). A funnel plot was used for the visualization of heterogeneity in the causal estimates from different genetic instruments. The “leave-one-out” test was used to identify whether a single SNP strongly affected the causal estimate. Additionally, we also applied Latent Heritable Confounder Mendelian randomization (LHC-MR) as a sensitivity analysis24. This method appropriately uses genome-wide genetic markers to estimate bidirectional causal effects, direct heritability, confounder effect and population stratification while accounting for sample overlap. LHC-MR extends the standard two-sample MR by modeling a latent heritable confounder that affects both the exposure and outcome traits. It can differentiate SNPs based on their co-association to a pair of traits. Thus, the unbiased bidirectional causal effect between these two traits is estimated simultaneously as well as the confounder effect on each trait. All MR analyses were performed using R (version 4.0.3). P < 0.05 was considered statistically significant.
Ethics approval
All methods were carried out following relevant guidelines and regulations. This study design was approved by the human research ethics committee in the Shanxi Medical University School and Hospital of Stomatology (2022SLL029).
Consent to participate
The present study is based on summary-level data that have been made publicly available. In all original studies, participate consent had been obtained.
Results
Causal effects of IBD and its subtypes on periodontitis
Character of selected SNPs
The first MR assumption was fulfilled through the quality control techniques. We identified 187, 162, and 125 independent SNPs associated with IBD, CD, and UC with P < 5 × 10−6, respectively. No palindromic genetic variant was seen among these SNPs. Among them, 179 SNPs for IBD, 157 SNPs for CD, and 119 SNPs for UC were available in the summary statistic of periodontitis. All SNPs were independent in LD clumping. The F-statistic values for all these SNPs were all larger than 10, suggesting that the weak instrument bias cannot substantially affect the causal estimates. The search in the PhenoScanner database further showed that 4 IBD SNPs, 9 CD SNPs, and 6 UC SNPs were associated with previously reported confounders, and they were thus excluded from subsequent analyses. Finally, 175 SNPs for IBD, 148 SNPs for CD, and 113 SNPs for UC were incorporated into the MR and sensitive analyses. Detailed information about these SNPs is shown in Tables S2–S4.
MR and sensitivity analysis
In the IVW analysis, the fundamental analysis, no statistically significant causal effects of genetically predicted IBD as a whole and its subtypes, CD and UC, on periodontitis were noted [IBD: odds ratio (OR) = 1.003, 95% CI = 0.977–1.030, P = 0.808; CD: OR = 1.023, 95% CI = 1.000–1.047, P = 0.051; UC: OR = 0.988, 95% CI = 0.958–1.019, P = 0.448] (Table 1, Fig. 2). The applied complementary MR approaches, the weighted median and MR Egger regression method, confirmed the results with point estimates slightly above and below 1 for the hypotheses—IBD affecting periodontitis—with highly overlapping CIs (Table 1, Fig. 2). The violations of the second and third assumptions were not found after statistical analysis. No heterogeneity was noted between the individual SNP of IBD and its subtypes in the heterogeneity analysis (IBD: P value of IVW method: 0.863, P value of MR Egger method: 0.858; CD: P value of IVW method: 0.514, P value of MR Egger method: 0.498; UC: P value of IVW method: 0.415, P value of MR Egger method: 0.405) (Table 2). No horizontal pleiotropy was detected in the Egger intercept test (IBD: intercept = 0.003, P = 0.439; CD: intercept = 0.002, P = 0.553; UC: intercept = 0.005, P = 0.428) (Table 3). The MR pleiotropy residual sum and outlier test outcomes showed that there were no horizontal pleiotropic outliers to distort the causality estimate. The “leave-one-out” test showed no outliers among these included SNPs. All sensitive analyses supported that there is no causal effect of genetically predicted IBD, UC, and CD on periodontitis. Figure 2 and Figs. S1–S3 show the funnel plots and leave-one-out analysis plots. LHC-MR also demonstrated no causal effect of IBD as a whole and its subtypes on periodontitis (IBD-Periodontitis: SE = 0.039, P = 0.858; UC-Periodontitis: SE = 0.071, P = 0.393; CD-Periodontitis: SE = 0.033, P = 0.470).
Causal effects of periodontitis on IBD and its subtypes
Character of selected SNPs
In the reverse direction, considering that no SNP was associated with periodontitis at the level of P < 5 × 10−8, the genetic instruments with P < 5 × 10−6 were selected. We identified 8 independent SNPs associated with periodontitis with P < 5 × 10−6. One of the genetic variants was palindromic and thus removed. The rest 7 SNPs were all available in the summary statistic of IBD and its subtypes. All these SNPs were independent in the LD clumping; their F-statistic values were all larger than 10. Another SNP was excluded after searching the PhenoScanner database. Finally, six SNPs remained as IVs of periodontitis; Tables S5–S7 present detailed information of these SNPs.
MR and sensitivity analysis
In the fundamental analysis, the IVW analysis, no statistically significant causal effects of genetically predicted periodontitis on IBD as a whole, CD and UC were found [periodontitis on IBD: odds ratio (OR) = 0.989, 95% CI = 0.898–1.088, P = 0.813; CD: OR = 0.945, 95% CI = 0.830–1.077, P = 0.398; UC: OR = 1.040, 95% CI = 0.928–1.165, P = 0.504] (Table 1, Fig. 3). The weighted median and MR Egger regression method showed the same results with point estimates slightly above and below 1 with highly overlapping CIs (Table 1, Fig. 3). No violations of the second and third assumptions were found after statistical analysis. No heterogeneity was noted between the individual SNP of periodontitis in the heterogeneity analysis (IBD: P value of IVW method: 0.128, P value of MR Egger method: 0.178; CD: P value of IVW method: 0.070, P value of MR Egger method: 0.135; UC: P value of IVW method: 0.214, P value of MR Egger method: 0.188) (Table 2). No horizontal pleiotropy was detected in the Egger intercept test (IBD: intercept = − 0.016, P = 0.298; CD: intercept = − 0.024, P = 0.250; UC: intercept = − 0.013, P = 0.478) (Table 3). The MR pleiotropy residual sum and outlier test outcomes demonstrated that there were no horizontal pleiotropic outliers. The “leave-one-out” test did not identify any leverage point with high influence. All sensitive analyses supported that there is no causal effect of genetically predicted periodontitis on IBD, UC, and CD. The funnel plots and leave-one-out analysis plots are shown in Fig. 3 and Figs. S4–S6. LHC-MR also demonstrated no causal effect of periodontitis on IBD as a whole and its subtypes (Periodontitis-IBD: SE = 0.358, P = 0.682; Periodontitis-UC: SE = 0.203, P = 0.944; Periodontitis-CD: SE = 0.319, P = 0.846).
Discussion
This two-sample MR among people of European descent revealed that the genetic liability for IBD, CD, and UC did not affect the risk of periodontitis or vice versa. Subsequent sensitivity analyses did not detect any horizontal pleiotropy and heterogeneity, further revealing the reliability of these results.
The association between periodontitis and IBD has been shown but with insufficient causal evidence. A large-scale cross-sectional study reported that Chinese patients with IBD have a higher prevalence, severity, and extent of periodontitis than controls9; similar results were noted in several case–control studies25,26. However, a gap exists between causality and association. In concept, cross-sectional cannot infer causal associations because of the absence of chronological sequence and so do case control studies mainly because of the high occurrence recall bias. Although meta-analyses evidence is now available, there is still remarkable heterogeneity in the methods and results of the study of the association between the two diseases27,28,29. Furthermore, these meta-analyses mainly included case–control studies, making weak causal inferences. Population-based studies with a cohort design provide an approach to infer causality as the onset of the exposure can be observed to happen before or after the outcome of interest. A large-scale cohort study reported an increased prevalence of subsequent periodontitis among patients with CD30. Notably, this study was conducted among Asian individuals instead of European individuals, and the population stratification effect makes it hard to conclude that its outcome contradicts our results31. Additionally, confounders can hinder cohort studies and make the causal association of these two illnesses unpredictable. For example, patients with periodontitis or IBD commonly suffer from numerous comorbidities; the common pathways of these comorbidities potentially contribute to the association between periodontitis and IBD32,33. MR does provide an effective strategy to decipher causality while significantly weakening the interference of confounder. It is reasonable to believe our present MR study was more sufficiently powered to assess the causality between periodontitis and IBD. In conclusion, the available studies demonstrated weak evidence on the causality between periodontitis and IBD in the European population, whereas our MR, serving as a robust alternative, indicated the absence of causality between IBD and periodontitis.
The hypothesized mechanism underlying the association of IBD and periodontitis remains elusive; the available evidence is insufficient to contradict the results of this two-sample MR. A current hypothesis of the association suggested an oral-gut interaction depending on the immune cell activation and microbiome to synergistically promote inflammation in genetically susceptible individuals34. However, no study on humans has conclusively proved this hypothesis yet. Some proposed pathways, such as calprotectin, are not specific and shared by several chronic inflammatory conditions with neutrophil infiltration30. These shared inflammatory pathways of these systemic diseases may confound the causal association and lead to discrepancies in outcomes between the previous observational studies and the present MR. Moreover, it is still unclear whether inflammatory processes in various peripheral tissues, such as periodontitis and IBD, are interlinked via inflammatory adaptations of the bone marrow, a central hub35. In summary, the causality between periodontitis and IBD has not been conclusively demonstrated at mechanism level; there is no reason to reject the results of this MR.
Interestingly, a recently published MR regarding IBD and periodontitis shows some different results to this study36. It concludes that IBD as a whole and UC have causal effect on periodontitis. In the reverse direction, periodontitis has an association with IBD and CD. This difference might root from the selection of datasets. Its GWAS of IBD and periodontitis both differ from ours in the aspect of sample size and population source. Specifically, the GWAS of IBD we use is a collective one, not only totally include GWAS used in the published MR, but also have additional new GWAS data18. The GWAS summary statistics of periodontitis we apply have five times more cases than their. Furthermore, both the two MR target European population. In their study, a single population source (Finland) of GWAS is employed to represent European population, but our GWAS of periodontitis has multiple population source (UK, U.S., Germany and Sweden). A detailed comparation of the GWAS summary statistics in the two studies is presented as a table (Table S8). Sample size and representativeness are key factors in determining the statistic power of MR outcomes13. Our GWAS, with larger sample size and multiple population sources can increase the statistic power of MR. In terms of the analysis methods, both studies apply almost the same methods including IVW, MR pleiotropy residual sum, outlier test, Weighted median and MR Egger regression. However, to further strengthen our analysis, we use LHC-MR as an additional sensitivity analyses24. This method appropriately uses genome-wide genetic markers to estimate bidirectional causal effects, direct heritability, confounder effect and population stratification while accounting for sample overlap. LHC-MR extends the standard two-sample MR by modeling a latent heritable confounder that affects both the exposure and outcome traits. It can differentiate SNPs based on their co-association to a pair of traits. Thus, the unbiased bidirectional causal effect between these two traits is estimated simultaneously as well as the confounder effect on each trait. According to LHC-MR, no bidirectional causal association exists in both directions, which further ensure the reliability of our result, make our outcome less prone to weak instrument bias and endow this MR with substantially high statistical power. In summary, the selection of GWAS leads to the difference of the two MR results in a large extent and more new GWAS and novel MR approaches are awaited for more authentic association between IBD and periodontitis.
The present study has four main strengths. First, in this two-sample MR, large-scale GWAS datasets from two independent populations of European descent were used to ensure sufficient statistical power for inferring a causal association and minimizing the effect of population stratification. Second, the high degree of consistency from various applied complementary MR approaches proved the stability of our study. Third, multiple sensitivity analyses showed no heterogeneity and pleiotropy and further supported our confidence in the established associations. Fourth, with the advantage of the MR study design, reverse causality could be eliminated and as did most potential confounding effects.
Nonetheless, there are several limitations to this study. First, the results were only drawn from the European population; thus, caution should be exerted when extrapolating the results to other ethnicities and races. Future MR studies need to incorporate additional genetic instruments thereby increasing the proportion of IBD and periodontitis explained. Second, given any genetic variant only explains a proportion of phenotypic variance, future MR studies should include larger samples and more valid IVs to make a robust causal inference. Third, these MR findings only reflected the change in outcomes due to a genetically predisposed (lifetime) status of exposures; the short-term effect of exposures on outcomes was unknown, highlighting the discrepancy between the results of this MR and the previous observational studies.
In conclusion, our two-sample MR study did not show a causal effect of genetically proxied IBD as an exposure on the development of periodontitis as an outcome using the summary statistics data within individuals of European ancestry. Similarly, in the reverse inference, limited evidence supported a causal role of genetic liability for periodontitis on IBD. Our results can provide a reference for future large-scale RCTs and might predict the results of these RCTs. However, caution should be exerted when it comes to clinical relevance and further studies are needed to clarify the relationship between IBD and periodontitis.
Data availability
The periodontitis summary statistic data are available at https://data.bris.ac.uk/data/dataset/. The IBD, CD, and UC summary statistic data can be accessed at https://www.ebi.ac.uk/gwas/studies/GCST004131, https://www.ebi.ac.uk/gwas/studies/GCST004132, https://www.ebi.ac.uk/gwas/studies/GCST004133.
Abbreviations
- CD:
-
Crohn’s disease
- CI:
-
Confidence intervals
- GLIDE:
-
Gene-Lifestyle Interactions in Dental Endpoints
- GWAS:
-
Genome-wide association study
- IBD:
-
Inflammatory bowel disease
- IV:
-
Instrumental variable
- LD:
-
Linkage disequilibrium
- MR:
-
Mendelian randomization
- OR:
-
Odds ratio
- RCT:
-
Randomized controlled trial
- SNP:
-
Single nucleotide polymorphism
- STROBE:
-
Strengthening the Reporting of Observational Studies in Epidemiology
- UC:
-
Ulcerative colitis
References
Baumgart, D. C. & Sandborn, W. J. Inflammatory bowel disease: Clinical aspects and established and evolving therapies. Lancet 369, 1641–1657 (2007).
Xavier, R. J. & Podolsky, D. K. Unravelling the pathogenesis of inflammatory bowel disease. Nature 448, 427–434 (2007).
Ni, J., Wu, G. D., Albenberg, L. & Tomov, V. T. Gut microbiota and IBD: Causation or correlation?. Nat. Rev. Gastroenterol. Hepatol. 14, 573–584 (2017).
Sn, P. et al. Inflammatory bowel disease and oral health: Systematic review and a meta-analysis. J. Clin. Periodontol. 44, 382–393 (2017).
Slots, J. Periodontitis: Facts, fallacies and the future. Periodontol 2000 75, 7–23 (2017).
Papapanou, P. N. et al. Periodontitis: Consensus report of workgroup 2 of the 2017 world workshop on the classification of periodontal and peri-implant diseases and conditions. J. Clin. Periodontol. 45 Suppl 20, S162–S170 (2018).
Hajishengallis, G., Chavakis, T. & Lambris, J. D. Current understanding of periodontal disease pathogenesis and targets for host-modulation therapy. Periodontol 2000 84, 14–34 (2020).
Shaddox, L. M., Morford, L. A. & Nibali, L. Periodontal health and disease: The contribution of genetics. Periodontology 2000(85), 165–181 (2021).
Zhang, L. et al. Increased risks of dental caries and periodontal disease in Chinese patients with inflammatory bowel disease. Int. Dent. J. 70, 227–236 (2020).
Habashneh, R. A., Khader, Y. S., Alhumouz, M. K., Jadallah, K. & Ajlouni, Y. The association between inflammatory bowel disease and periodontitis among Jordanians: A case–control study. J. Periodontal. Res. 47, 293–298 (2012).
Lorenzo-Pouso, A. I., Castelo-Baz, P., Rodriguez-Zorrilla, S., Pérez-Sayáns, M. & Vega, P. Association between periodontal disease and inflammatory bowel disease: A systematic review and meta-analysis. Acta Odontol. Scand. 79, 344–353 (2021).
Davey Smith, G., Paternoster, L. & Relton, C. When will mendelian randomization become relevant for clinical practice and public health?. JAMA 317, 589–591 (2017).
Emdin, C. A., Khera, A. V. & Kathiresan, S. Mendelian randomization. JAMA 318, 1925–1926 (2017).
Minelli, C. et al. The use of two-sample methods for Mendelian randomization analyses on single large datasets. Int. J. Epidemiol. 50, 1651–1659 (2021).
Labrecque, J. & Swanson, S. A. Understanding the assumptions underlying instrumental variable analyses: A brief review of falsification strategies and related tools. Curr. Epidemiol. Rep. 5, 214–220 (2018).
Skrivankova, V. W. et al. Strengthening the reporting of observational studies in epidemiology using Mendelian randomization: The STROBE-MR statement. JAMA 326, 1614–1621 (2021).
Shungin, D. et al. Genome-wide analysis of dental caries and periodontitis combining clinical and self-reported data. Nat. Commun. 10, 2773 (2019).
de Lange, K. M. et al. Genome-wide association study implicates immune activation of multiple integrin genes in inflammatory bowel disease. Nat. Genet. 49, 256–261 (2017).
Burgess, S., Butterworth, A. & Thompson, S. G. Mendelian randomization analysis with multiple genetic variants using summarized data. Genet. Epidemiol. 37, 658–665 (2013).
Bowden, J., Davey Smith, G., Haycock, P. C. & Burgess, S. Consistent estimation in Mendelian randomization with some invalid instruments using a weighted median estimator. Genet. Epidemiol. 40, 304–314 (2016).
Bowden, J., Davey Smith, G. & Burgess, S. Mendelian randomization with invalid instruments: Effect estimation and bias detection through Egger regression. Int. J. Epidemiol. 44, 512–525 (2015).
Hartwig, F. P., Davey Smith, G. & Bowden, J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int. J. Epidemiol. 46, 1985–1998 (2017).
Verbanck, M., Chen, C.-Y., Neale, B. & Do, R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat. Genet. 50, 693–698 (2018).
Darrous, L., Mounier, N. & Kutalik, Z. Simultaneous estimation of bi-directional causal effects and heritable confounding from GWAS summary statistics. Nat. Commun. 12, 7274 (2021).
Habashneh, R. A., Khader, Y. S., Alhumouz, M. K., Jadallah, K. & Ajlouni, Y. The association between inflammatory bowel disease and periodontitis among Jordanians: A case-control study. J. Periodontal. Res. 47, 293–298 (2012).
Vavricka, S. R. et al. Periodontitis and gingivitis in inflammatory bowel disease: A case-control study. Inflamm. Bowel Dis. 19, 2768–2777 (2013).
Nijakowski, K., Gruszczyński, D. & Surdacka, A. Oral health status in patients with inflammatory bowel diseases: A systematic review. Int. J. Environ. Res. Public Health 18, 11521 (2021).
Papageorgiou, S. N. et al. Inflammatory bowel disease and oral health: Systematic review and a meta-analysis. J. Clin. Periodontol. 44, 382–393 (2017).
She, Y.-Y. et al. Periodontitis and inflammatory bowel disease: A meta-analysis. BMC Oral Health 20, 67 (2020).
Chi, Y.-C. et al. Increased risk of periodontitis among patients with Crohn’s disease: A population-based matched-cohort study. Int. J. Colorectal Dis. 33, 1437–1444 (2018).
Lawlor, D. A., Harbord, R. M., Sterne, J. A. C., Timpson, N. & Davey Smith, G. Mendelian randomization: Using genes as instruments for making causal inferences in epidemiology. Stat. Med. 27, 1133–1163 (2008).
Hajishengallis, G. & Chavakis, T. Local and systemic mechanisms linking periodontal disease and inflammatory comorbidities. Nat. Rev. Immunol. 21, 426–440 (2021).
Olpin, J. D. et al. Beyond the bowel: Extraintestinal manifestations of inflammatory bowel disease. Radiographics 37, 1135–1160 (2017).
Newman, K. L. & Kamada, N. Pathogenic associations between oral and gastrointestinal diseases. Trends Mol. Med. https://doi.org/10.1016/j.molmed.2022.05.006 (2022).
Chavakis, T., Mitroulis, I. & Hajishengallis, G. Hematopoietic progenitor cells as integrative hubs for adaptation to and fine-tuning of inflammation. Nat. Immunol. 20, 802–811 (2019).
Wang, Z. et al. Association between inflammatory bowel disease and periodontitis: A bidirectional two-sample Mendelian randomization study. J. Clin. Periodontol. 50, 736–743 (2023).
Acknowledgements
The National Natural Science Foundation of China (Grant No. 82001075), the Natural Science Foundation of the Science and Technology Department of Shanxi Province (Grant No. 202103021224232), and the Science and Technology Innovation Foundation of Higher Education Institutions of Shanxi Province (Grant No. 2022L167) are gratefully acknowledged. Dr. Haworth and Dr. de Lange are acknowledged for providing summary statistics data of periodontitis and IBD.
Funding
This study was supported by the National Natural Science Foundation of China (Grant No. 82001075), the Natural Science Foundation of the Science and Technology Department of Shanxi Province (Grant No. 202103021224232), Shanxi Provincial Science and Technology Department Regional Cooperation and Exchange Project (Grant No. 202204041101026), the Basic Research Program of Shanxi Province (Grant No. 202203021212367) and the Science and Technology Innovation Project of Higher Education Institutions of Shanxi Province (Grant No. 2022L167).
Author information
Authors and Affiliations
Contributions
Y.Y., D.H., and F.Y. contributed to the study design; D.W., C.H., and M.C. contributed to data collection and analysis; Y.Y. wrote the first draft of the manuscript; D.H. and F.Y. revised and refined the manuscript; Y.Y., D.W., C.H., Q.W., and Y.L. contributed to data interpretation; All authors approved the final draft and accepted the decision to submit the manuscript for publication; D.H. and F.Y. contributed to funding acquisition.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher's note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Yu, F., Yang, Y., Wu, D. et al. Deciphering genetic causality between inflammatory bowel disease and periodontitis through bi-directional two-sample Mendelian randomization. Sci Rep 13, 18620 (2023). https://doi.org/10.1038/s41598-023-45527-z
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-023-45527-z
- Springer Nature Limited