Systematic Review and Meta-analysis on Association Between Air Pollutants and Hippocampal Volume from Magnetic Resonance Imaging in Adults

BACKGROUND AND AIM: The neurotoxicity of air pollutants has been actively investigated in recent years, and there is increasing epidemiological evidence suggesting that air pollution can adversely affect the central nervous system. In neurodegenerative disease, one important diagnostic biomarker is volume reduction in a key brain structure, the hippocampus, as assessed with neuroimaging techniques. Few epidemiological articles investigated the association of hippocampal volume with air pollution, with inconsistent results. In this paper, we aimed to estimate such association through a systematic review and meta-analysis. METHODS: Scopus and PubMed literature databases were searched through 31 March 2021. Eligibility criteria were: 1) assessment of air pollutant levels; 2) assessment of hippocampal volume through structural magnetic resonance imaging; 3) reporting of correlation coefficients along with standard error. We performed a meta-analysis using a random-effects model. RESULTS:We retrieved four studies using linear regression models to evaluate the possible effect of air pollutants on hippocampal volume of adult populations. The investigated air pollutants were nitrogen dioxide (NO2) and particulate matter, with diameter ≤2.5µm (PM2.5) and ≤10μm (PM10). All studies used a similar methodology based on standard spatial coordinates of images, and considered intracranial volume as a covariate. We found that hippocampal volume was inversely associated with PM2.5 concentration (β regression coefficient -7.90, 95% confidence interval (CI) -14.60 to -1.20) and more slightly with PM10 concentrations (β -1.70, 95%CI -3.96 to 0.57), whereas no association with NO2 concentrations emerged (β -0.29, 95%CI -1.13 to 0.55). CONCLUSIONS:Our results suggest that PM2.5 and less clearly PM10 have an adverse effect on hippocampal volume, a phenomenon associated with cognitive decline and increased risk of neurodegenerative disease. These epidemiologic findings appear to be biologically plausible especially for PM2.5, also taking into account some studies carried out in animals. KEYWORDS: Neurotoxicity, Neuroimaging, Magnetic Resonance Imaging, Hippocampus, Meta-analysis, Nitrogen dioxid