Genetically predicted obstructive sleep apnea is causally associated with an increased risk for periodontitis

Abstract

Background: Although obstructive sleep apnea (OSA) and periodontitis are associated, whether this association is causative is uncertain.

Methods: We conducted a bidirectional Mendelian randomization (MR) analysis using data from publically accessible genome-wide association studies. The single-nucleotide polymorphisms (SNPs) for OSA were derived from 16,761 cases and 201,194 controls. The pooled data of periodontitis association involved up to 17,353 individuals. Disease-associated single-nucleotide polymorphisms were selected as an instrumental variable at the genome-wide significance level (p < 5.0 × 10^{- 6}). Subsequently, the causal effects were estimated using three different methods: inverse variance weighting (IVW), MR-Egger, and weighted median. Then, these causal estimates were expressed as dominance ratios [odds ratio (OR)].

Results: The MR analysis revealed that genetically determined OSA promotes the development of periodontitis [ IVW OR = 1.117, 95% confidence interval (CI) = 1.001-1.246, p = 0.048). Furthermore, no causal effect of genetically predicted periodontitis on OSA was noted in the reverse MR analysis (IVW OR = 1, 95% CI: 0.95-1.06, p = 0.87). The trend in results from the MR-Egger regression and weighted median (WM) was consistent with that in results from the IVW method. The robustness of the results was confirmed by the sensitivity analysis.

Conclusions: In summary, the results of our MR investigation suggest an association between OSA and periodontitis, proposing that early screening and treatment of OSA is beneficial for the prevention and prognosis of periodontitis.

Keywords: Causality; Mendelian randomization; Obstructive sleep apnea; Periodontitis.

Fig. 1

Diagram of Mendelian randomization (MR) study design. (a) The evaluation of the causal effect of OSA on periodontitis. (b) The evaluation of the causal effect of periodontitis on OSA. (1) Each genetic variation (SNP) is linked to exposure (disease or phenotypic features); (2) SNPs are not associated with unmeasured confounders; and (3) SNPs can only affect the outcome via exposure; they cannot affect the outcome through other routes. OSA, obstructive sleep apnea; SNP, single nucleotide polymorphism

Fig. 2

Scatter plot depicts the causal estimates of the effect of OSA on periodontitis. Each point in the scatter plot represents an SNP. The effect of the same SNP on exposure is placed on the horizontal axis, and the effect on outcome is placed on the vertical axis. The vertical and horizontal lines show the 95% confidence interval (CI) for each SNP. At this point, the slope of the solid line in the plot is each Mendelian randomization (MR) estimate. OSA, obstructive sleep apnea; SNP, single nucleotide polymorphism

Fig. 3

The funnel plot (a) and leave-one-out plot (b) in the sensitivity analysis of the periodontitis-OSA relationship are presented. The funnel plot assesses the possible heterogeneity in the estimates. Each point indicates the inverse standard error corresponding to the individual causal estimates. The mean causal effect of all combinations of IVs (βIV) is indicated on the X-axis by inverse variance weighting (solid line) and the Mendelian randomization (MR)-Egger method (dashed line). The Y-axis indicates the inverse standard error of the estimated causal effect for each single nucleotide polymorphism (IVs). B The leave-one-out plot present how the causal estimates (point with horizontal circle) for the effect of OSA on periodontitis were influenced by the removal of a single variant. The bars indicate the confidence interval of MR estimates. OSA, obstructive sleep apnea; SNP, single nucleotide polymorphism