Causal Relationship Between Brain Functional Networks and Sleep Disorders: A Mendelian Randomization Study.

BACKGROUND

Dysfunction of brain resting-state functional networks has been widely reported in sleep disorders, but their causal relationships remain controversial.

OBJECTIVE

To investigate causal relationships between 191 resting-state functional magnetic resonance imaging (rsfMRI) phenotypes and eight sleep disorders using bidirectional two-sample Mendelian randomization (MR), with multivariable MR (MVMR) adjusting for addictive behaviors.

METHODS

Genome-wide association study (GWAS) data from UK Biobank and FinnGen were analyzed. Instrumental variables (IVs) were selected adhering to MR assumptions. Causal estimates were primarily derived via inverse-variance weighted (IVW), with sensitivity analyses using MR-Egger, weighted median, weighted mode, simple mode, and MR Pleiotropy Residual Sum and Outlier (MR-PRESSO).

RESULTS

We found increased functional connectivity within the salience and default mode networks in insular, cingulate, and frontal regions​​ (OR = 0.94, 95% CI: 0.92-0.97, pFDR = 0.042) was causally associated with a 6% reduced risk of daytime napping. For obstructive sleep apnea (OSA), ​​elevated functional connectivity within the default mode and central executive networks localized to precuneus/occipital regions​​ (OR = 1.16, 95% CI: 1.08-1.25, pFDR = 0.042) and ​​elevated functional connectivity within the default mode and central executive networks localized to parietal/temporal regions​​ (OR = 1.18, 95% CI: 1.09-1.27, pFDR = 0.016) having a causal effect, associated with a 16% and 18% elevated risk of OSA respectively. MVMR analysis revealed that coffee intake did not confound associations between rsfMRI phenotypes and sleep disorders, while smoking, alcohol consumption, tea intake, and cannabis use may act as confounding factors affecting these links.

CONCLUSION

This study provides robust evidence for causal associations between specific rsfMRI phenotype signatures and sleep disorders. It highlights brain functional networks as potential targets for non-invasive interventions, particularly within the context of addictive behaviors.