Genetic evidence supports a causal relationship between air pollution and brain imaging-derived phenotypes

Abstract

Background: Observational studies have reported associations between air pollutants and brain imaging-derived phenotypes (IDPs); however, whether this relationship is causal remains uncertain. Methods: We conducted bidirectional two-sample Mendelian randomization (MR) analyses to explore the causal relationships between 5 types of air pollutants (N=423,796 to 456,380 individuals) and 587 reliable IDPs (N=33,224 individuals). Two-step MR was also conducted to assess whether the identified effects are mediated through the modulation of circulating cytokines (N=8293). Results: We found genetic evidence supporting the association of nitrogen oxides (NOx) with mean intra-cellular volume fraction (ICVF) in the left uncinate fasciculus (IVW β=-0.42, 95 % CI −0.62 to −0.23, P=1.51×10−5) and mean fractional anisotropy (FA) in the left uncinate fasciculus (IVW β=-0.42, 95 % CI −0.62 to −0.21, P=4.89×10−5). In further two-step MR analyses, we did not find evidence that genetic predictions of any circulating cytokines mediated the association between NOx and IDPs. Conclusion: This study provides evidence for the association between air pollutants and brain IDPs, emphasizing the importance of controlling air pollution to improve brain health.

Keywords

• Mendelian randomization
• Air pollution
• Brain health
• Magnetic resonance imaging