In this large MR study with European participants, we performed a two-sample bidirectional MR analysis between four oral diseases with IBD and its two major subtypes. We found a strong association between genetically predicted lichen planus and increased risk of IBD and its major subtypes CD and UC by meta-analysis pooling multiple database outcomes. Additionally, we also discovered that IBD and its major subtypes were positively correlated with the risk of several oral diseases, with genetically predicted IBD and UC being associated with an increased risk of periodontal disease, genetically predicted IBD and CD being associated with an increased risk of lichen planus, as well as genetically predicted IBD being related to an increased risk of oral ulcers.
Periodontal disease
By performing MR analysis using more comprehensive database information, we did not find evidence of a causal effect between periodontal disease and IBD, which was inconsistent with the previous findings of Wang et al(15). This was probably due to the predominance of databases in which invalid causality was calculated in the effect statistics, which reverted more to true and objective results in the calculated of causal effect sizes in larger samples, allowing a modest causal relationship between periodontal disease and IBD to disappear(15). As a study by Ichikawa et al. confirmed on a more comprehensive microbiota characterization level(11), unlike animal periodontitis-colitis models, the presence of early periodontitis did not further enhance intestinal colonization by oral bacteria in IBD patients; however, periodontitis may exacerbate the development of IBD disease. In our study, the GWAS aggregated data we included for IBD focused on the outcome occurring or not, and due to the limitations of the data we were unable to observe the effect of periodontitis on the disease process of IBD. Hence, the presence of periodontitis may not lead to the development of IBD, but there was an effect on disease activity and progression, yet there was currently a lack of GWAS data on the disease process in IBD and therefore no causal evidence could be established. However, it was significant that the causal effect of IBD and its subtype UC on periodontal disease was once again confirmed in our MR study, which was in accordance with the MR findings of Wang et al(15). A meta-analysis system that included 1297 patients assessed the impact of IBD on periodontal disease and oral health(10). The findings suggest that IBD is associated with a significantly increased risk of periodontal disease and was more pronounced in UC than in CD. Although the pathological mechanisms between patients with IBD and periodontal disease have not been fully investigated, the two diseases shared pathogenetic similarities, and perturbations of multiple factors were suggested as possible mechanisms for the interaction between IBD and periodontal disease(35). On the one hand, microbial influences were thought to be a key factor in the increased susceptibility to periodontal disease in patients with IBD(13, 36). Brito et al. found that patients with IBD had higher levels of bacteria at sites of subgingival inflammation compared to non-IBD patients(37). It is emphasized that the colonizing microbiota in the oral cavity of IBD patients may play a role in disease pathogenesis as a modulator of the host response. On the other hand, the inflammatory response is an important factor in the increased risk of periodontal disease in patients with IBD(13). The study of Figueredo et al. found that serum IL-18 concentrations were higher in patients with periodontitis who had CD or UC compared to controls(38), a trend that was more pronounced in patients with IBD who were in active disease(39, 40). Aleksandra Nielsen et al. reported increased salivary levels of IL-6 in patients with CD(41), and Szczeklik et al. showed consistent results and found significantly higher levels of IL-6, IL-1β and TNF-α in patients with active CD than inactive patients and controls(39). Although existing studies have presented meaningful results on the mechanisms of action between IBD and periodontal disease, further research is needed to understand the complex interactions between microbial gene expression, environmental and genetic factors in the immune inflammatory response and dysregulated microbiota to improve the oral and gastrointestinal health of patients(42).
Lichen planus
Our bidirectional MR analysis explores for the first time the interaction between lichen planus and IBD, which adds important evidence to previous studies. Lichen planus is a chronic inflammatory skin mucosal disease involving the oral mucosa and can present in six different clinical manifestations: reticular, erosive/ulcerative, maculopapular, papular or atrophic. The etiology is currently unknown, but is thought to be related to immune-mediated mechanisms involving dendritic cells and T cell(7, 14), the oral cavity may be the only area involved(43). The presence of lichen planus can lead to an increased risk of malignant transformation of the oral cavity, and it is considered to be one of the important symptoms of a precancerous state(44). Previous observational studies have shown that IBD was associated with the development of lichen planus(7, 14, 44), and this may be attributed to the side effects of the drugs that were used, such as sulfasalazine and diltiazem(7). However, the study conducted by Femiano et al. denied the effects of drug complications(45). After discontinuing treatment drugs, oral symptoms did not improve and instead became more widespread. The results of our MR study also appear to show consistency, when removing as much as possible the confounding factors present in observational studies, we found that IBD and its major subtypes CD and UC were significantly correlated with an increased risk of lichen planus. Interestingly, the development of lichen planus was also associated with an increased risk of IBD and CD, but studies and evidence in this direction were reported sparingly, which seemed to be related to the pathogenesis of lichen planus. T cell-mediated immune responses play an important role in the pathogenesis of oral lichen planus (OLP)(43), and T cells may migrate to the oral epithelium through the 'chance encounter' hypothesis(43)and the 'directed migration' hypothesis(46, 47). In patients with oral lichen planus, the lymphocytic infiltrate consists almost exclusively of T cells, and most T cells in the epithelium and near damaged basal keratinocytes are activated CD8 + lymphocytes(43). Activated CD8 + T cells can release chemokines that attract additional lymphocytes and other immune cells to the developing OLP lesion(46, 48). Meanwhile, Jamwal et al. showed that CD8 + T cells were important in relation to the development and progression of colitis(49), that both CD4 + and CD8 + circulating T cells were activated in patients with IBD, and that the transcriptional profile of activated circulating CD8 + T cells predicted a more active disease cours(50). Similarly, a significant increase in the number of activated CD8 + T lymphocytes can be found in the intestinal lamina propria of CD and UC patients, including pathogenic subsets and immunosuppressive subsets(51). Therefore, CD8 + T cells may play an important role in the pathway between lichen planus and IBD, but the causal relationship between the two and the possible mechanisms of transmission should be explored in future research.
Oral ulcer
In addition, with our bidirectional MR study, we found a significant causal effect between IBD and oral ulcers, which was consistent with the results of previous several studies(52–55). Oral ulcers are associated with a high risk of inflammation and bacterial infection(56), Previous studies have shown that oral ulcers were the most common oral manifestation in IBD patients. Compared to the control group and other oral problems, IBD patients had the highest incidence of oral ulcers (36%), and patients with more active diseases tend to experience more adverse oral symptoms(57). Dysbiosis and ectopic flora of the gut microbiota in patients with IBD may be an important route to opportunistic infections in the oral cavity. Altered microbiome leads to dysbiosis and secondary inflammatory responses in the oral cavity, leading to the onset and development of oral problems(8), inducing oral ulcers, halitosis and tongue problems. Moreover, salivary dysfunction may be another important reason for the decline in oral health. Saliva is important for the protection of bacteria and fungi, for the maintenance of teeth and oral mucosa, and for the transport of enzymes and nutrients(57). de Vries et al. showed that in patients with CD, one of the main subtypes of IBD(57), a high prevalence of oral ulcers and dry mouth were reported, especially in patients with active disease, which could be attributed to salivary dysfunction, mainly in the form of inadequate salivary secretion and changes in salivary composition (e.g. increased MUC5B output), making saliva more viscous and thus causing numerous oral problems. This study identifies for the first time a moderate causal effect between genetically predicted IBD and oral ulcers from the perspective of gene sequencing, independent of confounding by potential confounders such as socioeconomic factors, prescribing policies and reverse causality, providing more objective results for unbiased estimation of causal effects.
Dental caries
According to our results, there was no significant correlation between IBD and the incidence of dental caries through a bidirectional MR analysis of large aggregate-level GWAS database. This view seems to contradict most research views(58–60), but this did not mean that our research view is on their opposite side. Actually, several systematic evaluations indicated that the prevalence of dentine caries in IBD patients was high, especially in CD patients(7, 60, 61). A small sample pilot study (n = 15) showed an association between IBD and a high prevalence of dental caries, but this result was influenced by multiple factors(62), including IBD patients' eating habits and reduced saliva. First of all, in order to improve symptoms, IBD patients usually choose a low-fat, high carbohydrate and high sugar diet, which significantly increases the incidence rate of dental caries(60). In particular, IBD patients generally take a high carbohydrate diet, while UC patients have a high sugar intake(63). Secondly, the decrease of saliva and the change of oral microorganisms in IBD patients are also important reasons for the high incidence of dental caries(7). Saliva plays an important role in protecting bacteria and fungi, as well as maintaining the function of teeth and mucous membranes(57). Inadequate salivation changes the oral microbiota, increasing the risk of dental caries and periodontitis(64, 65), the severity of xerostomia in CD patients increased significantly. Thirdly, Oral hygiene also plays a role in the high prevalence of dental caries in IBD patients. It is reported that the dental plaque index of IBD patients is higher than that of non IBD patients(61). These factors leading to bias were difficult to control in observational studies, but we can get more objective results through MR genome analysis to reduce bias. The same findings are reflected in these studies(61, 66, 67).
There were some limitations in this study: First of all, our study population is all European, the ethnographic differences of ancestors may lead to different results. Therefore, further research requires verification of other ethnic groups. Secondly, the sample size of CD in the FinnGen study and the oral diseases (including periodontal disease, lichen planus and dental caries) was small, in order to enable more comprehensive results to be observed, we used P < 1 × 10− 5 for the extraction of IVs. As to reduce the bias introduced by weak IVs, we performed multiple sensitivity analyses and tests of heterogeneity, as well as calculated F-statistics for individual SNPs to enhance the reliability of the IVs. Thirdly, in the MR analysis of periodontal disease and IBD, there may be some overlap in the samples when using FinnGen study as an exposure-outcome database. Finally, subgroup analyses could not be performed because summary statistics were conducted, and analyses could not be stratified by gender and ethnicity(33, 68).