Abstract Rationale Wildfire-derived fine particulate matter (PM2.5) is associated with increased risk of acute respiratory exacerbations and hospitalizations. However, it remains unclear whether long-term exposure to wildfire-derived PM2.5 may also contribute to lung function decline and impaired respiratory health, particularly in the general population. Methods In the CARDIA study, a population-based cohort of adults enrolled from four centers in the United States (Oakland, Chicago, Birmingham, Minneapolis), five metrics capturing the frequency, duration, and intensity of long-term wildfire-derived PM2.5 exposure were estimated from 2006 to 2010. Daily mean PM2.5 levels at participants’ home addresses were estimated by the Stanford Environmental Change and Human Outcomes Lab using surface measurements, satellites, and machine learning. We examined the association between wildfire-derived PM2.5, measured as continuous variables and quartiles, and change in lung function and plasma biomarkers of impaired respiratory health. Spirometry measured in 2005/2006 and then in 2015/2016 determined incident obstruction, incident preserved ratio impaired spirometry (PRISm) and change in FEV1 and FVC percent predicted. Circulating protein levels from plasma collected in 2010/2011 were used to calculate the “respiratory susceptibility score,” a previously developed proteomic risk score associated with accelerated lung function decline and predictive of incident chronic lung disease, respiratory exacerbations and respiratory mortality. Multivariable logistic and linear regression models were adjusted for age, sex, race, education, BMI, smoking status, and smoking pack years at the 2005/2006 visit. Results Among 2,730 participants, the mean number of weeks between 2006-2010 with wildfire-derived PM2.5 5 μg/m³ was 4.8 (SD 3.3), the mean daily wildfire PM2.5 concentration was 2.4 μg/m3 (SD 1.4), and the mean daily wildfire PM2.5 during peak exposure week was 3.3 μg/m3 (SD 1.7), and each were independently associated with incident obstruction, incident PRISm, change in FEV1 and FVC, and the respiratory susceptibility score. (Table 1). When comparing participants with the lowest quartile of weeks with wildfire-derived PM2.5 5 μg/m³, those with the highest quartile of exposure had 74% higher odds of incident obstruction (OR 1.74 95% CI 1.09-2.74), 66% higher odds of incident PRISm (OR 1.66 1.05-2.59), and 0.19 point higher respiratory susceptibility score (β 0.19 0.08-0.30). Conclusions In a U.S. population-based cohort, greater frequency and intensity of long-term exposure to wildfire-derived PM2.5 are independently associated with incident impaired lung function and circulating proteins associated with the development of chronic lung disease and respiratory morbidity. This suggests that long-term exposure to wildfire smoke may increase the risk of chronic lung disease. This abstract is funded by: NHLBI, NCATS, ALA
Liu et al. (Fri,) studied this question.