Associations of source-specific PM2.5 and incident disability in older U.S. adults: the Health Retirement Study

Abstract

Keywords: Air Pollution, emission sources, Activities of Daily Living, PM2.5 Background and Aims: Late-life disability is a common endpoint of the chronic health conditions that affect people as they age. Although fine particulate matter (PM2.5) is associated with chronic conditions that contribute to disability, associations with disability itself are understudied, and the most toxic emission sources have yet to be identified. We evaluated the association between long-term exposure to PM2.5 from specific emission sources and disability. Methods: We used biennial survey data between 2006 and 2016 on Activities of Daily Living (ADL) from participants >65 years of the Health and Retirement Study. Using a spatiotemporal prediction model, we estimated 10-year PM2.5 concentrations at participants' residences and multiplied these values by spatially-resolved fractions of PM2.5 from sources including agriculture, energy production, industry, transportation, fires, and windblown dust, derived from a chemical transport model. We used Cox models to estimate the risk of incident ADL disability per 1-µg/m³ source-specific PM2.5, adjusting for individual- and area-level confounders, time, geography, and sampling weights. Results: Among 11,782 adults (72±6.8 years), 37% reported a new disability during follow-up. Overall, both total PM2.5 and source-specific PM2.5 were associated with increased incident of ADL disability. A 1-µg/m³ increment in total PM2.5 corresponded to a 2% (95% CI: 1.00, 1.04) greater hazard of ADL. In single-sector models, PM2.5 from each source was also positively associated with incident ADL. After adjusting for PM2.5 from other sources, associations with agriculture, non-road and road transportation, non-coal energy production, and windblown dust PM2.5 remained, whereas no associations were observed for the other sources. Conclusions: We found evidence that higher levels of PM2.5 from some but not all sources were associated with increased risk of disability, suggesting that selective interventions could be effective at improving health and function in late-life.