Genetic Evidence Prioritizes Neurocognitive Decline as a Causal Driver of Sleep Disturbances: A Multi-Omics Analysis Identifying Causal Genes and Therapeutic Targets.

To resolve the ambiguous causal relationship between sleep disturbances and neurodegenerative diseases such as Alzheimer's disease (AD), we conducted a multi-stage genetic and multi-omics investigation. Our large-scale bidirectional Mendelian randomization analysis identified a robust, asymmetrical pattern of genetic association, providing strong genetic evidence suggesting that liability for neurocognitive decline and AD is associated with sleep disturbances, with substantially weaker evidence for the reverse direction. To identify the underlying molecular drivers, a multi-omics Summary-data-based MR (SMR) analysis prioritized high-confidence causal genes, including YWHAZ, NT5C2, COX6B1, and CDK10. The predictive power of this gene signature was confirmed using machine learning models (ROC-AUC > 0.8), while functional validation through bulk and single-cell transcriptomics uncovered profound, cell-type-specific dysregulation in the AD brain, most notably opposing expression patterns between neurons and glial cells (e.g., YWHAZ was upregulated in excitatory neurons but downregulated in glia). Functional enrichment and network analyses implicated two core pathways-nucleotide metabolism centered on NT5C2 and synaptic function involving YWHAZ-and our investigation culminated in the identification of a promising therapeutic interaction, with molecular docking validating high-affinity binding between Ecdysterone and COX6B1 (docking score = -5.73 kcal/mol). Collectively, our findings strengthen the evidence that sleep disruption as a likely consequence of neurodegenerative processes and prioritize a set of validated, cell-type-specific gene targets within critical pathways, offering promising new avenues for therapeutic development.