Association of ambient air pollution and pesticide mixtures on respiratory inflammatory markers in agricultural communities

Abstract

Air pollution exposure is associated with adverse respiratory health outcomes. Evidence from occupational and community-based studies also suggests agricultural pesticides have negative health impacts on respiratory health. Although populations are exposed to multiple inhalation hazards simultaneously, multidomain mixtures (e.g. environmental and chemical pollutants of different classes) are rarely studied. We investigated the association of ambient air pollution-pesticide exposure mixtures with urinary leukotriene E4 (LTE4), a respiratory inflammation biomarker, for 75 participants in four Central California communities over two seasons. Exposures included three criteria air pollutants estimated via the Community Multiscale Air Quality model (fine particulate matter, ozone, and nitrogen dioxide) and urinary metabolites of organophosphate (OP) pesticides (total dialkyl phosphates (DAPs), total diethyl phosphates (DE), and total dimethyl phosphates (DM)). We implemented multiple linear regression models to examine associations in single pollutant models adjusted for age, sex, asthma status, occupational status, household member occupational status, temperature, and relative humidity, and evaluated whether associations changed seasonally. We then implemented Bayesian kernel machine regression (BKMR) to analyse these criteria air pollutants, DE, and DM as a mixture. Our multiple linear regression models indicated an interquartile range (IQR) increase in total DAPs was associated with an increase in urinary LTE4 in winter (β: 0.04, 95% CI: [0.01, 0.07]). Similarly, an IQR increase in total DM was associated with an increase in urinary LTE4 in winter (β:0.03, 95% CI: [0.004, 0.06]). Confidence intervals for all criteria air pollutant effect estimates included the null value. BKMR analysis revealed potential non-linear interactions between exposures in our air pollution-pesticide mixture, but all confidence intervals contained the null value. Our analysis demonstrated a positive association between OP pesticide metabolites and urinary LTE4 in a low asthma prevalence population and adds to the limited research on the joint effects of ambient air pollution and pesticides mixtures on respiratory health.

Keywords: ambient air pollution; bayesian kernel machine regression; central california; environmental mixtures; leukotrienes; organophosphate pesticides; respiratory health.

Figure 1.

Spatial variation in the concentrations of O₃, PM_2.5, and NO₂ in Fresno and Tulare Counties in January and June 2019, shown as the monthly averages based on NAAQS daily standards, as rasterized values estimated from the CMAQ model. Red points represent study community locations.

Figure 2.

Spearman correlations among DE and DM with total DAPs, air pollutants, temperature, and relative humidity.

Figure 3.

Univariate relationship between each environmental exposure and log-transformed LTE4, while holding all other exposures at their median value both in January and in June. These results were assessed using the BKMR model adjusted for age, sex, asthma diagnosis, current employment in agriculture, living with someone currently employed in agriculture, temperature, and relative humidity.

Figure 4.

Relationship between DAPs and urinary LTE4 z-scores at the 25th, 50th, and 75th percentiles of an air pollutant exposure in January.

Figure 5.

Relationship between DAPs and urinary LTE4 z-scores at the 25th, 50th, and 75th percentiles of an air pollutant exposure in June.