Associations of chemical composition and sources of PM2.5 with lung function of severe asthmatic adults in a low air pollution environment of urban Nagasaki, Japan

Previous studies have linked ambient PM2.5 to decreased pulmonary function, but the influence of specific chemical elements and emission sources on the severe asthmatic is not well understood. We examined the mass, chemical constituents, and sources of PM2.5 for short-term associations with the pulmonary function of adults with severe asthma in a low air pollution environment in urban Nagasaki, Japan. We recruited 35 asthmatic adults and obtained the daily record of morning peak expiratory flow (PEF) in spring 2014–2016. PM2.5 filters were extracted from an air quality monitoring station (178 days) and measured for 27 chemical elements. Source apportionment was performed using Positive Matrix Factorization (PMF). We fitted generalized linear model with generalized estimating equation (GEE) method to estimate changes in PEF (from personal monthly maximum) and odds of severe respiratory deterioration (first ≥ 15% PEF reduction within a 1-week interval) associated with mass, constituents, and sources of PM2.5, with adjustment for temperature and relative humidity. Constituent sulfate (SO42−) and PM2.5 from oil combustion and traffic were associated with reduced PEF. An interquartile range (IQR) increase in SO42− (3.7 μg/m3, average lags 0–1) was associated with a decrease of 0.38% (95% confidence interval = −0.75% to −0.001%). An IQR increase in oil combustion and traffic-sourced PM2.5 (2.64 μg/m3, lag 1) was associated with a decrease of 0.33% (−0.62% to −0.002%). We found a larger PEF decrease associated with PM2.5 from dust/soil on Asian Dust days. There was no evidence linking total mass and metals to reduced pulmonary function. The ventilatory capacity of adults with severe asthma is susceptible to specific constituents/sources of PM2.5 such as sulfate and oil combustion and traffic despite active self-management of asthma and low air pollution levels in the study location.