Obstructive Sleep Apnea Linked to Retinal Thinning Without Clear Genetic Causality

A comprehensive systematic review and meta-analysis of 62 observational studies has established that patients diagnosed with obstructive sleep apnea exhibit significantly thinner retinal nerve fiber layers compared to healthy individuals. Although clinical measurements demonstrated consistent structural degradation that worsened alongside illness severity, complementary bidirectional two-sample Mendelian randomization analyses failed to substantiate a direct causal relationship between the two conditions.

Prior physiological hypotheses suggested that the disorder may actively contribute to progressive retinal neurodegeneration through repetitive cycles of intermittent hypoxia and systemic vascular dysregulation.

To resolve these questions, investigators conducted a systematic search across eight primary academic databases from their inception through February 2026, supplemented by specialized registries including ClinicalTrials.gov, the Chinese Clinical Trial Registry, and Google Scholar. The observational meta-analysis was quantitatively synthesized using RevMan 5.4 and Stata 15.1 software, incorporating formal evaluations for subgroup contrasts, sensitivity analyses, and potential publication bias. To address potential confounding inherent in observational study designs, the authors concurrently designed a bidirectional two-sample Mendelian randomization framework utilizing genome-wide association study datasets from the GWAS Catalog for retinal parameters and the FinnGen R13 release for obstructive sleep apnea.

Across the 62 observational studies that met all inclusion criteria, the meta-analysis demonstrated that the average retinal nerve fiber layer thickness among patients with obstructive sleep apnea was significantly reduced compared to healthy control participants. The pooled analysis identified a mean difference of -3.56 micrometers (95% confidence interval: -4.98 to -2.15, P < 0.00001).

Further quadrant-based evaluations revealed that tissue thinning occurred across all four primary anatomical sectors of the eye, though the extent of loss varied by location. Significant thinning was documented in the temporal, nasal, superior, and inferior quadrants, showing mean differences of -1.57, -3.04, -3.53, and -4.66 micrometers, respectively, with all regional reductions reaching statistical significance at P < 0.001.

Stratification according to clinical disease severity uncovered distinct patterns of neuroretinal decline across stages of the disorder. In patients with mild obstructive sleep apnea, significant thinning was restricted entirely to the temporal quadrant, showing a mean difference of -1.09 micrometers (95% confidence interval: -2.03 to -0.15, P = 0.02), while average thickness and remaining quadrants showed no significant difference compared to healthy controls. By contrast, individuals diagnosed with moderate and severe obstructive sleep apnea exhibited widespread, statistically significant reductions across overall retinal nerve fiber layer thickness and throughout all four individual anatomical quadrants.

To evaluate potential sources of observational heterogeneity, the researchers conducted formal subgroup tests based on geographic region. These comparisons showed that geographic location produced a statistically significant difference exclusively within the superior quadrant analysis (P = 0.02). Conversely, no significant between-region divergence was identified for average retinal thickness, temporal thickness, nasal thickness, or inferior thickness, demonstrating substantial cross-population consistency for most measured ocular sectors.

Genetic causal inference was assessed using inverse variance weighted methods alongside weighted median, MR-Egger, simple mode, and weighted mode estimators. In the forward direction, genetically predicted retinal nerve fiber layer thickness showed no causal effect on sleep apnea risk under the primary inverse variance weighted model (odds ratio = 1.002, 95% confidence interval: 0.988 to 1.017, P = 0.749). In the reverse analysis, testing the causal effect of sleep apnea on retinal thickness, the primary inverse variance weighted estimate was also not statistically significant (beta = 0.159, standard error = 0.237, P = 0.503), despite nominally significant positive estimates emerging in the weighted median and weighted mode tests that indicated conflicting behavior among estimators.

Subject of Research: Observational association and genetic causal inference between obstructive sleep apnea and retinal nerve fiber layer thickness

Article Title: Association and causal inference between obstructive sleep apnea and retinal nerve fiber layer thickness: a meta analysis and bidirectional Mendelian randomization study

Article References: Xu, Y., Li, J., Yan, L., Gou, L., Lei, X., Chen, Q., Su, A., He, C., Zhu, L., Guo, H., & Liu, J. (2026). Association and causal inference between obstructive sleep apnea and retinal nerve fiber layer thickness: a meta analysis and bidirectional Mendelian randomization study. BMC Ophthalmology. https://doi.org/10.1186/s12886-026-05453-w

Image Credits: AI Generated

DOI: 10.1186/s12886-026-05453-w

Keywords: Obstructive Sleep Apnea, Retinal Nerve Fiber Layer, RNFL Thickness, Retinal Neurodegeneration, Mendelian Randomization, Meta-Analysis, Intermittent Hypoxia, Vascular Dysregulation, Inverse Variance Weighted, FinnGen R13, GWAS Catalog, Temporal Quadrant, Inferior Quadrant, Superior Quadrant, Nasal Quadrant, Disease Severity, Observational Studies, BMC Ophthalmology