AbstractBackgroundAmbient air pollutants are toxic to the vascular systems, and recent studies have shown the late‐life exposures may result in neurodegeneration and increased risks for AD. It is unclear if cerebrovascular damage and AD‐related neurodegeneration contribute to the atrophies of medial temporal lobe. Here we assessed whether AD related neurodegeneration and cerebrovascular disease mediated adverse effects of air pollution exposure on MTL atrophy.MethodParticipants included 624 community‐dwelling older women without dementia (mean age 77.4 +/‐ 3.4 years) who underwent two structural brain MRI scans (MRI‐1: 2005‐6; MRI‐2: 2010‐11) as part of the Women’s Health Initiative Memory Study. Regionalized universal kriging was used to estimate annual concentrations of PM2.5 (fine particulate matter) and NO2 at residential locations aggregated to 3‐year averages prior to MRI‐1. The outcome was 5‐year standardized change in the MTL volumes. Mediators included spatial pattern of neurodegeneration of AD (AD pattern similarity scores [ADPS], validated ADNI data) and a proxy of global cerebrovascular damage (small‐vessel ischemic disease [SVID]). Whether the associations between exposures on MTL atrophy were mediated by the initial level or concurrent change in ADPS or global SVID were examined by the structural equation models with latent‐difference scores. All effect paths were adjusted for sociodemographic, lifestyle, clinical characteristics, and intracranial volume.ResultLiving in locations with higher PM2.5 (per interquartile range [IQR] = 3.21µg/m3) and NO2 (per IQR = 6.72ppb) were both associated with greater atrophy of the MTL (βPM2.5 = ‐.27, 95% confidence interval [CI] = [‐.38,‐.16]; βNO2 = ‐.12,95%CI = [‐.22,‐.02]). PM2.5 was associated with increased ADPS (βPM2.5 = .02, 95%CI = [.01,.05]) over time, which partially mediated associations with MTL atrophy (26%, 95% CI = [10%,49%; Figure 1 Panel A). NO2 was positively associated with ADPS at MRI‐1 (βNO2 = .02, 95%CI = [.01,.04]), which partially mediated the association with MTL atrophy (14%, 95% CI = [1%,94%]; Figure 1 Panel B). Global SVID at MRI‐1 or change in SVID did not significantly mediate observed associations between exposures and MTL atrophy (Figure 1 Panels C and D).ConclusionOur study supports the mediating role of AD neurodegeneration contributing to MTL atrophy associated with late‐life exposures to air pollutants. Findings support the hypothesis that multiple neuropathological processes of brain aging are affected by air pollution neurotoxicity.