Causal relationships between genetically predicted particulate air pollutants and neurodegenerative diseases: A two-sample Mendelian randomization study

Abstract

Accumulating observational studies have linked particulate air pollutants to neurodegenerative diseases (NDDs). However, the causal links and the direction of their associations remain unclear. Therefore, we adopted a two-sample Mendelian randomization (TSMR) design using the GWAS-based genetic instruments of particulate air pollutants (PM2.5 and PM10) from the UK Biobank to explore their causal influence on four common neurodegenerative diseases. Estimates of causative relationships were generated by the Inverse variance weighted (IVW) method with multiple sensitive analyses. The heterogeneity and pleiotropy tests were additionally performed to verify whether our findings were robust. Genetically predicted PM2.5 and PM10 could elevate the occurrence of AD (odds ratio [OR] = 2.22, 95 % confidence interval [CI] 1.53–3.22, PIVW = 2.85×10−5, PFalse discovery rate [FDR] = 2.85×10−4 and OR = 2.41, 95 % CI: 1.26–4.60, PIVW = 0.008, PFDR =0.039, respectively). The results were robust in sensitive analysis. However, no evidence of causality was found for other NDDs. Our present study suggests that PM2.5 and PM10 have a detrimental effect on AD, which indicates that improving air quality to prevent AD may have pivotal public health implications.

Keywords

• PM2.5
• PM10
• Mendelian randomization
• Neurodegenerative diseases
• Causal inference
• Alzheimer’s disease