[A two-sample Mendelian randomization study on the association between temporomandibular disorder and insomnia].

Abstract

Objectives: This study aimed to investigate the association between temporomandibular disorder (TMD) and insomnia using a two-sample Mendelian randomization (MR) approach.

Methods: Bidirectional MR analyses of two samples, TMD (n=377 277) and insomnia (n=375 359), were performed using genome-wide association study statistics published in the FinnGen database. Instrumental variables were first screened, and then inverse variance weighting (IVW) and MR-Egger were used as the main-effect assessment methods. Weighted median, weighted mode, and simple mode served as supplementary methods. We used IVW and MR-Egger to test for heterogeneity, as well as MR-Egger intercepts to assess the single nucleotide polymorphism (SNP) potential level of multiplicity effects. Sensitivity analyses were conducted based on leave-one-out to identify potentially influential SNPs. All analyses were conducted by using the two-sample MR R package and were considered statistically significant when P<0.05.

Results: MR analysis showed the presence of TMD on insomnia (OR=1.089, 95%CI: 1.017-1.166, P=0.014). Meanwhile, no effect of insomnia on TMD (OR=0.996, 95%CI: 0.964-1.029, P=0.816) was found. The sensitivity-analysis showed that no heterogeneity existed (P>0.05), and the presence of horizontal pleiotropy was not detected (P>0.05). Leave-one-out sensitivity analysis showed no single SNP, which may affect the causal relation. All findings indicated that the causal relationship between TMD and insomnia was not significantly affected by any individual SNP and that IV did not bias the results.

Conclusions: Results of MR analyses showed that TMD is a risk factor for insomnia, whereas insomnia is not a risk factor for TMD.