Examination of the causal association between lung function and Alzheimer’s Disease: a Mendelian Randomization study

Abstract

Introduction: Observational studies have reported associations between low lung function and dementia risk but cannot test causality of this relationship. If causal, interventions to maintain lung function with ageing could reduce the burden of dementia. Aims and Objectives: As RCTs are impossible, we used Mendelian randomisation (MR, analysis that uses genetic variants as instrumental variables) to investigate whether lower lung function causes Alzheimer’s disease (AD). Methods: We performed two sample MR studies using single nucleotide polymorphisms found in two case-control genome wide association studies (GWAS) for low lung function (Wain et al. 2017 & UKBB by Neale Lab 2017) and one case-control GWAS for AD. We selected genetic instruments consisting of 58 – 177 independent signals for FEV1 and/or FEV1/FVC on 17,008 patients with AD and 37,154 controls (Lambert et al Nat Genet 2013 Dec;45(12):14252-8). In each study the individual causal effect estimates for each SNP were combined using the inverse-variance weighted (IVW) method. Results: The causal effect estimates on AD from the IVW method were Odd Ratios (95% Confidence intervals) 1.04 (0.91, 1.19) and 1.02 (0.48, 2.18) per unit decrease in FEV1 or FEV1/FVC with p-values≥0.58. We found no strong evidence (MR-Egger p-values≥0.69 and PWM pvalues≥0.12). Conclusion: We found very weak evidence for a causal effect of lung function on risk of AD. Previous observational associations may have been confounded.