The association between long-term PM2.5 exposure and late-life amyloid burden in the Atherosclerosis Risk in Communities (ARIC) study cohort

Background and Aim: Mechanistic models suggest that particulate matter ≤ 2.5 um (PM2.5) may promote Alzheimer’s disease, which is characterized by brain amyloid accumulation. This hypothesis has rarely been explored in epidemiologic studies, despite studies showing associations between long-term PM2.5 exposure and late-life cognitive impairment. We estimated the association between long-term PM2.5 exposure and late-life brain amyloid deposition in the Atherosclerosis Risk in Communities (ARIC) study cohort. Methods: We used a chemical transport model with data fusion to estimate PM2.5 concentrations (ug/m3) in 36-, 12-, 4-, and 1-km grid cells in ARIC study areas. We linked concentrations to geocoded participant addresses and calculated mean PM2.5 exposures from 2000 to 2007. We estimated amyloid deposition using florbetapir amyloid positron emission tomography (PET) scans in 346 participants with normal cognition or mild cognitive impairment in 2011-2014. We defined amyloid positivity as a global cortical standardized uptake value ratio (SUVR) ≥ the sample median of 1.2. We used logistic regression models to quantify the association between amyloid positivity and PM2.5 exposure after adjusting for potential confounders. We explored effect measure modification by APOE e4 allele status and tested whether effect estimates were consistent using alternate PM2.5 exposure methods. Results: Our analytic sample included 279 participants. At the time of amyloid-PET scans, their mean age was 78 years, 56% were female, 42% were Black, and 26% had mild cognitive impairment. After adjusting for age, sex, education, and race-study center, we found no significant association between brain amyloid positivity and long-term mean PM2.5 exposure. There was no evidence of effect measure modification by APOE e4 allele status. Results were consistent when we used alternate PM2.5 estimation methods. Conclusions: PM2.5 may induce neurotoxic effects through non-amyloid, potentially vascular pathways, though we note the small sample size may have made us underpowered to detect a significant association.