Causal association and molecular mechanisms linking insomnia to late- onset Alzheimer’s disease: an integrated Mendelian randomization and transcriptomic analysis

Background Insomnia is consistently associated with Alzheimer’s disease (AD) in observational studies, but the causal relationship and the underlying molecular mechanisms remain unclear. This study aimed to investigate the causal relationship between insomnia and late-onset AD and to identify molecular pathways and candidate genes that may contribute to this association by integrating genetic causal inference with transcriptomics. Methods Two-sample Mendelian randomization (MR) was performed using 28 single-nucleotide polymorphisms as instruments, with insomnia genome-wide association study summary statistics from UK Biobank (336,965 participants of European ancestry) and late-onset AD data from FinnGen (11,359 cases and 508,613 controls). Inverse-variance weighting was the primary method. Robustness was assessed using MR-Egger, weighted median, simple mode, and weighted mode estimates, together with pleiotropy, heterogeneity, and leave-one-out analyses. GEO datasets GSE122063 and GSE110226 were integrated for differential expression, weighted gene co-expression network, protein interaction, and functional enrichment analyses. Lasso regression and support vector machine recursive feature elimination were used to identify key candidate genes, whose discrimination was assessed using receiver operating characteristic curves. Results Inverse-variance weighting associated genetically predicted insomnia with higher odds of late-onset AD (OR = 3.36, 95% CI 1.02–10.99; P = 0.045). Other MR methods yielded directionally consistent but imprecise estimates. The MR-Egger intercept test found no evidence of directional pleiotropy (P = 0.939), and Cochran’s Q test was not statistically significant (P = 0.062). No individual instrument dominated the leave-one-out results. Transcriptomic analysis identified 1,616 differentially expressed genes and eight co-expression modules. MEpink, MEblack, and MEbrown showed the strongest associations with AD. Their intersection with 2,330 insomnia-related genes yielded 35 candidates, enriched mainly in neuroinflammation, cell adhesion, and lipid metabolism. Both machine learning algorithms selected seven key candidates. All seven were differentially expressed between AD and controls in the combined dataset, with individual AUCs above 0.75. Conclusion Genetic liability to insomnia may be associated with an increased risk of late-onset AD. Neuroinflammation and lipid metabolism may link the two conditions, and seven candidate genes warrant further study. Given the borderline MR result, wide confidence interval, tissue heterogeneity, and absence of independent external validation, these findings require confirmation in larger multi-ancestry cohorts, independent transcriptomic datasets, and experimental studies.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-19
DOI
https://doi.org/10.21203/rs.3.rs-11643513/v1
Primary Topic
Dementia and Cognitive Impairment Research
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Causal association and molecular mechanisms linking insomnia to late- onset Alzheimer’s disease: an integrated Mendelian randomization and transcriptomic analysis

Yuanyuan Jiang, Xiaoni Zhang, Xiaomeng Zhao, Shan Qin et al.
Research Square
Dementia and Cognitive Impairment Research
preprint

Causal association and molecular mechanisms linking insomnia to late- onset Alzheimer’s disease: an integrated Mendelian randomization and transcriptomic analysis

Yuanyuan Jiang, Xiaoni Zhang, Xiaomeng Zhao, Shan Qin, Liu Chengyong, Lu Lu, Wenzhong Wu
preprint en

Abstract

Background Insomnia is consistently associated with Alzheimer’s disease (AD) in observational studies, but the causal relationship and the underlying molecular mechanisms remain unclear. This study aimed to investigate the causal relationship between insomnia and late-onset AD and to identify molecular pathways and candidate genes that may contribute to this association by integrating genetic causal inference with transcriptomics. Methods Two-sample Mendelian randomization (MR) was performed using 28 single-nucleotide polymorphisms as instruments, with insomnia genome-wide association study summary statistics from UK Biobank (336,965 participants of European ancestry) and late-onset AD data from FinnGen (11,359 cases and 508,613 controls). Inverse-variance weighting was the primary method. Robustness was assessed using MR-Egger, weighted median, simple mode, and weighted mode estimates, together with pleiotropy, heterogeneity, and leave-one-out analyses. GEO datasets GSE122063 and GSE110226 were integrated for differential expression, weighted gene co-expression network, protein interaction, and functional enrichment analyses. Lasso regression and support vector machine recursive feature elimination were used to identify key candidate genes, whose discrimination was assessed using receiver operating characteristic curves. Results Inverse-variance weighting associated genetically predicted insomnia with higher odds of late-onset AD (OR = 3.36, 95% CI 1.02–10.99; P = 0.045). Other MR methods yielded directionally consistent but imprecise estimates. The MR-Egger intercept test found no evidence of directional pleiotropy (P = 0.939), and Cochran’s Q test was not statistically significant (P = 0.062). No individual instrument dominated the leave-one-out results. Transcriptomic analysis identified 1,616 differentially expressed genes and eight co-expression modules. MEpink, MEblack, and MEbrown showed the strongest associations with AD. Their intersection with 2,330 insomnia-related genes yielded 35 candidates, enriched mainly in neuroinflammation, cell adhesion, and lipid metabolism. Both machine learning algorithms selected seven key candidates. All seven were differentially expressed between AD and controls in the combined dataset, with individual AUCs above 0.75. Conclusion Genetic liability to insomnia may be associated with an increased risk of late-onset AD. Neuroinflammation and lipid metabolism may link the two conditions, and seven candidate genes warrant further study. Given the borderline MR result, wide confidence interval, tissue heterogeneity, and absence of independent external validation, these findings require confirmation in larger multi-ancestry cohorts, independent transcriptomic datasets, and experimental studies.

Research Square
Nanjing University of Chinese Medicine (CN), Nanjing Medical University (CN), Nanjing University (CN)
Dementia and Cognitive Impairment Research
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon