The causal association between temporomandibular joint disorders and eating disorders: a bidirectional two-sample mendelian randomization study

Background The association between temporomandibular joint disorders (TMD) and eating disorders (ED) is controversial owing to the observational design of the relevant studies. This study aimed to explore the potential causal association between TMD and ED utilizing the Mendelian randomization (MR) approach. Materials and methods A bidirectional two-sample MR was conducted. The Single nucleotide polymorphisms (SNPs) related to the exposures (TMD and ED) were extracted from publicly available genome-wide association studies at the level of P = 5 x 10(-5). The analyses included Inverse-variance weighted (IVW), weighted median, MR-Egger, simple mode, and weighted mode methods. Heterogeneity was tested using Cochran's Q-test. Horizontal pleiotropy was assessed using the MR-Egger intercept. The sensitivity and reliability of the results were assessed by the Leave-one-out analysis. Moreover, the radial IVW model, radial MR-Egger model, MR-PRESSO, and MR-RAPS were used for causal estimates, the validity of the SNPs, and the detection of outliers. P < 0.05 was considered significant. Results The MR analysis revealed a causal effect of TMD on ED (IVW, odds ratio (OR) = 1.09, 95% confidence interval (CI) = 1.02-1.17, P = 0.014; MR-PRESSO, OR = 1.09, 95%CI = 1.03-1.15, P = 0.005; IVW(Mod.2nd), b = 0.08, se = 0.02, P = 0.003). However, there was no causal effect of ED on TMD (P > 0.05). Neither heterogeneity nor horizontal pleiotropy was found, and the results were confirmed to be robust. Conclusions This study provides evidence that generically predicted TMD can lead to ED. However, the opposite is not true. Therefore, addressing TMD must be emphasized as an approach to preventing and treating associated ED.