Abstract 4361590: Atrial Fibrillation as a Causal Risk Factor for Dementia: Pooled Mendelian Randomization Analysis of 2.6 Million Individuals from Large-Scale Genome-wide Association Studies

Background: Atrial fibrillation (AF) has been linked to dementia in prior observational studies. Whether this reflects a causal relationship remains to be a topic for debate, as both are strongly associated with age and various comorbid conditions. Prior mendelian randomization (MR) studies have been performed to establish causality by using genetic variants as instruments for AF. We conducted a meta-analysis of MR studies to assess the association between genetically predicted AF and dementia, including Alzheimer’s disease (AD) and other subtypes, which has never been attempted before. Methods: We systematically identified published two-sample MR analyses of AF (using different genome-wide association study instruments, e.g. AFGen consortium meta-analysis, FinnGen) in relation to dementia outcomes. All outcome summary statistics were drawn from large-scale consortia (e.g. International Genomics of Alzheimer's Project (IGAP) for AD) and cohorts of predominantly European ancestry. We extracted odds ratios (ORs) for each dementia subtype and combined estimates using random-effects meta-analysis. Heterogeneity was evaluated with the I 2 statistics. P<0.05 was considered statistically significant. Results: This meta-analysis (5 studies) pooled data from approximately 211,000 AF cases and over 2.6 million controls across large-scale GWAS consortia (AFGen, FinnGen, UK Biobank, IGAP, PGC-ALZ, and others), predominantly in individuals of European ancestry (Table 2). Genetically predicted AF showed a significant positive association with vascular dementia (pooled OR 1.21; 95%CI 1.07–1.37; heterogeneity I 2 44.2%). In contrast, there was no significant association between AF and AD (pooled OR 1.00, 95%CI 0.98–1.02; P=0.47). Similarly, no causal effect was observed on Lewy body dementia (LBD) or frontotemporal dementia (FTD) (Figure 1). All MR estimates were directionally consistent across studies and there was no evidence of pleiotropy in sensitivity analyses as well. Conclusion: In our meta-analysis of MR studies, a genetic predisposition to AF was associated with higher odds of vascular dementia, however no clear causal link was evident for combined dementia, AD, other dementia subtypes. These findings suggest AF itself may contribute to vascular dementia. Further research is warranted to understand pathophysiology of AF leading to vascular dementia and to inform preventive strategies aimed at reducing cognitive morbidity and improving long-term outcomes in AF patients.