Assessing the causal association between celiac disease and Alzheimer disease and frontotemporal dementia: A bidirectional Mendelian randomization approach

This study aimed to investigate the bidirectional causal relationship between celiac disease (CD) and the risk of Alzheimer disease (AD) or frontotemporal dementia (FTD) using Mendelian randomization (MR), in order to clarify prior inconsistent findings. We analyzed summary-level genome-wide association study (GWAS) data for CD, AD, and FTD. Single-nucleotide polymorphisms (SNPs) strongly associated with each condition were selected as genetic instruments. MR analysis was conducted in 2 directions: from CD to AD/FTD and from AD/FTD to CD. Mendelian Randomization Pleiotropy RESidual Sum and Outlier (MR-PRESSO) was used to detect and correct for pleiotropy, and Cochran Q assessed heterogeneity. Leave-one-out and Mendelian Randomization-Egger (MR-Egger) regression sensitivity analyses were performed to evaluate robustness. No evidence of a causal effect was found between CD and either AD or FTD in either direction (P > .05). Similarly, genetic liability to AD or FTD did not increase the risk of CD. Sensitivity analyses supported the robustness of the results, showing no pleiotropy or heterogeneity. Our findings suggest that CD is not causally linked to the development of AD or FTD. While shared genetic factors or comorbidities may exist, the association is likely noncausal, and other mechanisms of cognitive decline in CD patients warrant further study.