Relationship between long-term exposure to sources of particulate matter, ambient air pollution and cognitive...

Long-term exposure to particulate matter with a diameter of less than 2.5 μm (PM_2.5) is associated with cognitive decline, but it is not clear whether the potential effects of PM_2.5 from different emission sources vary. This study aims to assess the association between PM_2.5 from different emission sources and cognitive function in older adults in Northern Ireland. Using data of the Northern Ireland Cohort of Longitudinal Ageing from wave 1 (n = 3440) and wave 2 (n = 5408). Long-term exposure to PM_2.5 (5-year and 10-year average) was estimated using the UK Pollution Mapping approach within the 1-km residential circular buffers, while PM_2.5 source fraction was calculated using the atmospheric chemistry-transport model. Cognitive function was assessed using the Mini-Mental State Examination (MMSE) scores. The association between PM_2.5 and MMSE scores was estimated using multilevel linear regression models, and the Quantile-based g-computation approach was employed to decompose the potential effect of different emission sources of PM_2.5. The 5-year average PM_2.5 levels were negatively associated with MMSE scores (coef. = -0.07; SE = 0.02), with no evidence of association with 10-year average levels (coef. = -0.05; SE = 0.02). Furthermore, 5-year average PM_2.5 from agriculture, wildfires and road traffic were negatively associated with MMSE scores, while 10-year average PM_2.5 from agriculture was also negatively associated with MMSE outcomes. Moreover, industry and agricultural sources provide the greatest contributions to the potential effect of PM_2.5. Long-term exposure to PM_2.5, especially from agricultural, road traffic, and wildfire, may be harmful to cognition. Therefore, policy should aim to regulate emissions from high-risk sources.