Evaluating the Causal Effect of Type 2 Diabetes on Alzheimer's Disease Using Large-Scale Genetic Data

D Liu,A Baranova,F Zhang

DOI: https://doi.org/10.14283/jpad.2024.148

Abstract:Background: Alzheimer's disease (AD) has a high comorbidity with type 2 diabetes (T2D). However, there is still some controversy over whether T2D has a causal impact on AD at present. Objectives: We aimed to reveal whether T2D has a causal effect on AD using large-scale genetic data. Methods: Firstly, we performed a primary two-sample Mendelian randomization (MR) analysis to assess the potential causal effects of T2D on AD. For this analysis, we used the largest available genome-wide association studies (GWAS) T2D (T2D1, including 80,154 cases and 853,816 controls) and AD (AD1, including 111,326 cases and 677,663 controls) datasets. Additionally, we performed a validation MR analysis using two largely overlapping-sample datasets from FinnGen, including T2D (T2D2, including 57,698 cases and 308,252 controls) and AD (AD2, including 13,393 cases and 363,884 controls). In all MR analyses, the inverse variance-weighted method was used as the primary analysis method, supplemented by the weighted-median and MR-Egger techniques. Results: In the primary analysis, we found that T2D was not associated with the risk of AD (OR: 0.98, CI: 0.95-1.01, P=0.241). Similarly, no significant association was detected in the validation MR analysis (OR: 0.97, CI: 0.64-1.47, P=0.884). Conclusion: Our findings provide robust evidence that T2D does not have a causal impact on AD. Future studies need to further explore the effect of T2D on the non-AD components of the dementia phenotype.