Modifiable Risk Factors for Alzheimer's Disease: a Mendelian Randomization perspective

Abstract

Background: Despite extensive efforts, modifiable risk factors for Alzheimer s disease (AD) are still poorly understood. The long prodromal phase of AD, the presence of confounders, competing risk factors, and the risk of reverse causation make the detection of causal risk factors particularly challenging. To address these challenges, we used Mendelian Randomization to investigate the role of modifiable risk factors in AD. Method(s): We considered 30 genetically instrumentable risk factors as exposures. We used the AD GWAS from the IGAP consortium as the primary outcome, which ensured thorough classification and age-matching of cases and controls. Using this data, we performed extensive bidirectional univariable and multivariable Mendelian Randomization (MR) analyses. Result(s): Whilst we could validate the well-known protective effect of intelligence and the risk-increasing role of LDL cholesterol, the effect of other risk factors did not align with the previous literature. Namely, we found evidence that a higher BMI reduces the risk for AD, mediated by lean mass. Furthermore, higher systolic blood pressure (SBP) robustly reduced the risk for AD. This effect was neither driven by competing risk factors nor by reverse causation, but we could replicate the risk-increasing effect of higher SBP on traits related to vascular dementia (VaD). Conclusion(s): Beyond well-established risk factors for AD (such as intelligence and LDL cholesterol), we also found that higher SBP and higher lean mass had a robust protective effect on AD. Our findings demonstrate the importance of carefully assessing distinct components of exposures to understand how they relate to different forms of dementia.