This is the first study to investigate the causal relationship between RAS and NOMAP using MR analysis. Our results suggested that RAS was not significantly causally related to NOMAP. Among all the causal analyses, only BNFM was found to be related to RAS by the IVW method, but it may be due to horizontal pleiotropy, as the subsequent MR-Egger regression analysis suggested. The existence of horizontal pleiotropy shows that genetic variation might influence a different trait through several distinct routes, contradicting assumptions (2) and (3) of MR analysis, rendering the causal impact estimates derived inaccurate [23]. For the other outcomes, the Cochran Q-test, MR-Egger regression, and MR-PRESSO did not reveal any heterogeneity or directional pleiotropy.

As research on the relationship between inflammation and tumors has become more in-depth, many types of tumors have been found to be associated with chronic inflammation. Such as Crohn’s disease, ulcerative colitis, and inflammatory bowel disease, which may progress to colorectal cancer, and bronchitis, which may result in lung cancer [30, 31]. On the other hand, research indicates that while those who have inflammatory bowel illness have a higher risk of acquiring colorectal cancer, only around 2% of those who have the condition had inflammatory bowel disease prior to the cancer emerging [32]. There is no exact evidence for the inevitable connection between the inflammation and tumors, as certain damage can be allowed to continue via DNA repair mechanisms, which can lead to mutagenesis but not always malignancy [33].

Inflammation was initially emphasized for its important role in defense against pathogens and for its contribution to tissue repair, regeneration, and remodeling [34]. Subtle forms of inflammation can play a crucial role in regulating tissue homeostasis [35]. The normal inflammatory response typically initiates when an infection or injury damages epithelial tissue, triggering the activation of myeloid cells [36]. These cells then produce inflammatory cytokines, activate innate and adaptive immunity, eliminate pathogens, and stimulate the proliferation of epithelial cells to close the barrier dysfunction that leads to pathogen translocation and repair other injuries caused by the stimulus [36]. As a result of the concerted effort, the damaged epithelial tissue returned to its normal homeostatic state [37]. However, if the original disruption of epithelial homeostasis is triggered by an oncogenic event, immunity will not repair the damage, and increased inflammation and cytokine-driven proliferation will promote tumour development rather than restore normal epithelial homeostasis [37].

Therefore, identification of the causative factor is the key to determining whether the inflammation is carcinogenic or not. The etiology and pathophysiology of RAS are quite complex. It has been proven to be influenced by a variety of variables, including nutritional inadequacy, microbial flora imbalance, stress, unhealthy lifestyle choices, medicines, allergens, psychological issues, anemia, immunological disorders, genetic predisposition, etc [5]. While certain causative factors such as tobacco, alcohol, and betel nuts, identified in our research, are recognised risk factors for NOMAP linked to the development of RAS, the majority of other causative factors have not been definitively proven to be carcinogenic [38,39,40]. If there is no other oncogenic event, the RAS is unlikely to be a carcinogenic inflammation.

The ambiguity surrounding the etiology of RAS, the vulnerability of observational studies to confounding factors, reverse causality, and various biases, along with the challenges of conducting experimental studies, complicate the potential of traditional research methodologies to establish a causal relationship between RAS and NOMAP.

Our work utilized the MR method and confirmed at the genetic level that no causal association between RAS and NOMAP was identified, contributing to addressing them dialectically and the formulation of clinical strategies for them.

There are some major strengths in our MR analysis. First, in contrast to observational research, the inclusion of genetic variations as IVs lessens the possibility of common confounding variables and reverse causation. Second, in order to reliably research the causal relationship between RAS and NOMAP, we employed SNP-exposure and SNP-outcome estimations from studies with the greatest sample sizes to date (varying from 287916 to 461106 individuals). Third, the GWAS dataset we used was based primarily on populations of European ancestry, which minimized the effect of population stratification [41]. Fourth, our conclusion is based on a comprehensive analysis involving eight oropharyngeal tumor characteristics using a variety of causal estimation models, heterogeneity tests, and sensitivity tests, which can effectively reduce the occurrence of various biases and ensure the validity and stability of the results.

The limitations of our study are as follows. First, RAS in the UKB was diagnosed from a questionnaire instead of a medical evaluation. This restriction is essential due to the unavailability of clinical oral examination data and the brief, intermittent occurrence of mouth ulcers, which frequently renders them invisible during clinical examinations even to those afflicted [18]. As with any data obtained from questionnaires, this may result in some misclassification, exemplified by our study’s inability to differentiate between various types of mouth ulcers. Nevertheless, we anticipate that our results primarily represent the causal assoiation between RAS and NOMAP, as various groups have corroborated the majority of variations from UKB, including three variants specific to RAS, the predominant form of oral ulcer [42]. Second, we could not completely eliminate the effects of horizontal pleiotropy, despite implementing a comprehensive series of measures to identify and mitigate aberrant variations. This is probably attributable to the intricate biological functions of numerous genetic variants. Third, due to the scarcity of GWAS data for non-white populations and the potential for population stratification arising from the amalgamation of GWAS data across multiple races, our analysis only used data from white individuals, excluding data from other racial groups [41]. When applying our results to populations outside Europe, care must be taken, since numerous environmental factors may significantly impact RAS and NOMAP. Fourth, we did not take into account sex-specific effects, and the incidence of RAS and NOMAP may differ by sex due to hormonal changes. Finally, to corroborate the findings, more substantial sample numbers and more sophisticated techniques are required.