Workflow for Comparison of Chemical and Biological Metrics of Filter Collected PM2.5

Abstract

There is limited understanding of adverse health effect associations with chemical constituents of fine particulate matter (PM_2.5) as well as the underlying mechanisms. We outlined a workflow to assess metrics, beyond concentration, using household and personal PM_2.5 filter samples collected in India as a proof of concept for future large-scale studies. Oxidative potential, chemical composition (polycyclic aromatic hydrocarbons and elements), and bioactivity (developmental exposures in zebrafish) were determined. Significant differences were observed in all metrics between personal and household PM_2.5 samples. This work established methods to characterize multiple metrics of PM_2.5 to ultimately support the identification of more health-relevant metrics than concentration.

Keywords: Household air pollution; Indoor PM2.5; PAH diagnostic ratios; PM2.5 composition; oxidative potential; zebrafish.

Figure 1:

Experimental Design. Schematic of the preparation of PM_2.5 following collection from kitchen or personal monitors from households in Kheri, India. PM_2.5 was used for chemical and biological response data.

Figure 2:

PM_2.5 Concentrations from Households with Kitchen and Personal Monitors. Concentrations (μg/m³) are reported for each individual filter collected from either a kitchen or personal monitor worn by a female participant in six households (A-F). Dashed lines indicate the mean concentration for the particular monitor type across all sampled households.

Figure 3:

Oxidative potential for PM_2.5 from kitchen and personal monitors with individual household and pooled samples. Data are reported as mean ± standard deviation of DTT consumption (nmol) per minute per μg PM_2.5. All analyses were run in triplicate and statistical significance of a p-value ≤ 0.05 was determined by two-way ANOVA with * denoting significant differences between monitor types from individual households (A-F) and a,b denoting significance between pooled and individual samples for kitchen or personal monitors, respectively.

Figure 4:

PAH and element concentrations in PM_2.5 from kitchen or 309 personal monitors. Following pooling of household samples by monitor type (each group comprised of 6 filter extracts), aliquots of the PM_2.5 kitchen and personal samples were analyzed for PAHs (n=114, a) via GC-MS and elements (n=14, b) via ICP-OES. Data is reported as concentration ± standard error mean (SEM). Data was blank corrected (laboratory and filter blank) and normalized by μg of PM_2.5 with all samples run in triplicate. Statistical significance between monitor types was determined by one-way ANOVA, with p ≤ 0.05 designated as * and p ≤ 0.001 as **.

Figure 5:

Detected constituents in kitchen and personal PM_2.5 samples. Following pooling of household samples by monitor type (each group comprised of 6 filter extracts), aliquots of the PM_2.5 kitchen and personal samples were analyzed for PAHs (n=114) and elements (n=14). A) Total number of detected compounds are reported for classes of PAHs and elements. B) Venn diagram of overlap of detected compounds between the monitor types.